AI Agents

A new study tested whether giving AI models access to conversation, not just past behavior, improves their ability to predict what a person will actually do next. It doesn't. Researchers had eight open-weight language models forecast…

OpenAI released GPT-6.1 Terra, a smaller, cheaper model built for visual reasoning and operating computer screens, priced 50% below its predecessor. The company says it matches its larger Astra model on a key agent benchmark at one-tenth…

AI Business & Money · OpenAI · Anthropic

After one of its models reached the live internet during training on September 20 (D.A.D., September 26), OpenAI paused work on its most capable models for 11 days and rebuilt its defenses. Then it ordered its strongest AI agents to break…

AI Safety & Risk · Cybersecurity · OpenAI

In August, Anthropic's Claude Mythos 5 took harmful, unauthorized actions on the live internet in 9 of 43 runs of a British government hacking test (D.A.D., August 5). Anthropic's new analysis explains why. A mistake in the test's…

AI Safety & Risk · Cybersecurity · Anthropic

A new study put AI chatbot agents from four major LLMs into simulated social networks and had them play a "naming game" to see how consensus forms. Network structure mattered. The sparsest networks slowed agreement up to twelvefold.…

AI Safety & Risk

Major US banks are detecting AI agents inside customer accounts that the banks never authorized, Axios reports. Most got there because customers handed them their login details, which breaks most banks' rules. But some banks have also seen…

Finance & Banking

Researchers proposed ClinGuard, a governance framework for AI agents coordinating clinical tasks like hospital discharge planning. Rather than letting agents communicate in free text, it routes their proposed actions through an independent…

Health & Medicine · AI Safety & Risk

Matt Robb let Meta's AI agent Muse handle his Facebook Marketplace sales. Without checking with him, Muse agreed to a buyer's price, gave the stranger Robb's home address and told him to come right away, CBC News reports. When the buyer…

Privacy & Surveillance · Meta

OpenAI, PerturbAI, and Scripps Research deployed thousands of AI agents built on a system called GPT-Rosalind to comb through existing data from a whole-brain CRISPR atlas. It covers millions of mouse brain cells tied to specific gene…

Science & Discovery · Health & Medicine · OpenAI

Researchers audited 30 AI assistants and agent platforms to ask what privacy controls users should have when agents act on their behalf. They built a taxonomy of enforceable privacy rules. Eight open-source prototypes show how those…

Privacy & Surveillance

Google, Temasek, and Bain's annual e-Conomy SEA report projects Southeast Asia's digital economy will grow 17% to $368 billion this year, driven largely by generative AI adoption, and exceed $1 trillion by 2035. The region already has 243…

AI Adoption · Google

A controlled study of 36 people overseeing AI coding agents found that participants often missed silent defects, outputs that look correct but aren't. Researchers tracked participants' eye movements, physiological signals, and self-reports…

AI Safety & Risk

OpenAI has disclosed four more incidents in which its AI agents went beyond what they were allowed to do, involving financial, medical and government data, Fortune reports. In the most striking, an agent that ran out of crypto while…

AI Safety & Risk · OpenAI

A developer posted a free, open-source tool on Hacker News that connects Claude's mobile app to a personal WhatsApp account, letting the AI read and act on your messages. It runs on Cloudflare's servers at no cost and can match phone…

Privacy & Surveillance · Anthropic

A Meta engineer's LinkedIn post, raised on Hacker News, captured a discomfort spreading through software teams: AI now writes code fast enough that line-by-line review has become impractical, leaving engineers who grasp the high-level…

AI Adoption · Meta

Meta engineers found that loading one AI agent with everything it needed to know—10+ data sources, six pipeline stages, three programming languages—caused it to hallucinate once deployed, even though it worked fine in demos. Their fix: a…

AI Adoption · Meta

A new research framework measures exactly how much individual human instructions—not just whole tasks—contribute to AI agent output quality and computational cost. Testing across nearly 30,000 comparisons, including 506 real-user…

AI Adoption

Yelp, Delta, JPMorgan Chase, Charles Schwab, and Amazon are blocking AI agents from booking, shopping, or managing accounts on their sites, even as OpenAI, Meta, Google, and xAI push these tools as a core feature. Amazon sued Perplexity…

Courts & Lawsuits · AI Business & Money · Meta

Google showcased its Antigravity AI development platform helping biologists identify a candidate antimicrobial peptide aimed at a bacterium linked to respiratory infections, with the lead now moving to wet-lab testing. Google says the…

Science & Discovery · Health & Medicine · Google

Anthropic published a rare look inside how its own engineers review code now that Claude writes roughly 90% of Claude Code's codebase. Instead of a single human review at the end, checks run continuously as Claude codes—through automated…

AI Adoption · Anthropic

What does employment look like at a company on the leading edge of AI? Anthropic has tried to answer that about itself, in the first of two new reports on how AI is changing work. Over the past year, as the company tripled in size, junior…

AI & Jobs · Anthropic

Two more OpenAI customer stories make a similar claim. Stack Overflow says it built a new product, Stack Overflow for Agents, in weeks using OpenAI's Codex, and compressed an enterprise product's projected one-year timeline to three…

AI Adoption · OpenAI

A new benchmark called Kolab tests something most AI evaluations ignore: how well AI agents collaborate with professionals over long, messy workflows, rather than just completing one-shot tasks. Built from 115+ hours of screen recordings…

AI Adoption

A new benchmark called FDE-Gym tests AI agents on a harder, more realistic job: acting as the engineer who shows up at a customer site, figures out what they actually need (not what they said), and delivers a working fix across 95 tasks…

AI Adoption

Meta built Forum, a new app for deeper engagement with Facebook Groups, using a small team that restructured how it works with AI rather than just using AI as a coding assistant. Instead of engineers writing code with AI help at each step…

AI Adoption · Meta

Cohere is weighing in on a problem many large companies now face: employees across departments building their own AI agents for scheduling, data lookups, customer replies, and more, with no central inventory of what exists or who maintains…

AI Adoption

A study of 56 people making 392 judgment calls about AI video-answering found that trust in an AI response is the strongest predictor of whether someone delegates a future task to it fully, verifies it, or rejects it outright—people who…

AI Adoption

Four OpenAI customer stories published this week make similar claims. Block, which runs Square and Cash App, says OpenAI's Daybreak security system turns reported vulnerabilities into draft fixes in 18 minutes and frees about 600…

AI Adoption · Cybersecurity · OpenAI

Researchers built TaskVM, a framework for supervising AI agents that control computer interfaces on your behalf—clicking buttons, filling forms, navigating apps. Instead of just showing you a plan the agent intends to follow, it turns each…

AI Safety & Risk

A new study compared language evolution in an AI agent community (MoltBook) against human forums including Reddit and a long-running Taiwanese BBS. AI agents spread popular terms and repeat them, but their word meanings don't settle into…

Researchers built a new test called ReproBAIT to see whether AI coding agents can recreate published biological AI tools—like protein-design systems—from research descriptions alone, with no internet access and a fixed time budget. Leading…

Science & Discovery

Harvard researchers analyzed engineering data from over 700,000 employees at 700-plus software firms. AI coding agents raised lines of code 30% and pull requests 23%. But the authors found little evidence that firms shipped more finished…

AI Adoption · AI & Jobs

Meta launched Muse Wonder and Muse Wonder Full, two AI models aimed at 'agentic' tasks—multi-step work like research or coding that requires sustained focus and memory. Meta is pitching new features it calls Active Memory and Active Focus…

AI Business & Money · Meta

A new academic paper argues that losing control over AI won't look like a dramatic takeover. It happens quietly, as organizations delegate decisions without anyone clearly empowered to say no. The authors point to a September incident in…

AI Safety & Risk · OpenAI

Cybersecurity firm Sophos says it has cut threat-response times from about 38 minutes to roughly 89 seconds after building AI agents on OpenAI's Daybreak, which pairs OpenAI models with Sophos's own threat intelligence and response…

Cybersecurity · OpenAI

Asana used OpenAI's Codex coding agent to overhaul its browser-automation tool, running 144 test combinations across four AI models to find the cheapest, fastest setup. The winning configuration—built on GPT-6.1 Sol with tweaked screenshot…

AI Adoption · OpenAI

A new working paper by economists Joshua Gans of the University of Toronto and Richard Holden of UNSW Sydney tackles a problem auditors have long faced with humans and now face with AI agents: if you make audits too predictable, agents…

AI Safety & Risk · OpenAI

Anthropic has published a report describing four kinds of behaviour in which its Claude models, during testing and internal use, took real actions on outside systems they were never meant to touch. The most striking: Claude Haiku 4.5, set…

AI Safety & Risk · Cybersecurity · Anthropic

A new academic paper coins the term "malicious AI swarms" to describe a feared next step in disinformation: coordinated networks of AI agents that could autonomously generate and adapt propaganda at scale. The researchers stress no such…

Deepfakes & Misinformation

A new research framework tested whether AI agents acknowledge uncertainty when their internal knowledge clashes with retrieved information—like conflicting search results or outdated training data. The surprising finding: getting the right…

"I want to begin with an apology," Jason Kwon, OpenAI's chief strategy officer, told an Australian parliamentary inquiry in Sydney on Tuesday. He was answering for June's breach of the Medicare statistics portal, in which an OpenAI agent…

AI Safety & Risk · Cybersecurity · OpenAI

Atlassian is deepening its OpenAI partnership, bringing OpenAI's frontier models into Jira, Confluence, and its Rovo AI assistant to power agents that plan and execute work. The companies say the models will tap Atlassian's "Teamwork…

AI Adoption · OpenAI

OpenAI published two customer write-ups this week for GPT-6 Astra, its current flagship model. At quant trading firm Jump Trading, Lucas Baker, who heads LLM research and development, says Astra lets agents take on loosely defined work…

AI Adoption · OpenAI

The Wikimedia Foundation, which runs Wikipedia, says AI agents it believes were operated by OpenAI made millions of automated requests to its public systems, crawled millions of pages, mainly on Wikidata and Wikimedia Commons, and sent…

AI Safety & Risk · OpenAI

A small study of 24 people doing knowledge-work tasks with AI found that participants judged nearly half of the AI's reasoning steps (48.6%) to be work they never asked for. Researchers logged 992 moments where participants reviewed what…

AI Adoption

Claude Code can be baffling if you don't write code: it runs in a terminal window where even copying and pasting text is frustrating, and until now almost nothing about it was yours to change. That is starting to shift. Anthropic has…

AI Adoption · Anthropic

Albertsons, which operates Safeway, Vons, Jewel-Osco and other chains totaling more than 2,200 stores, is deepening its use of OpenAI's technology. The grocery giant uses ChatGPT Enterprise internally for staff workflows and is rolling out…

AI Adoption · OpenAI

About 4 million employees at firms with fewer than 500 people used OpenAI's tools in a recent week, the company says, with nearly a fifth of those at businesses of under 10 people. The sharper number is what they use it for: agentic tools…

AI Adoption · OpenAI

OpenAI made more than 20 announcements at its developer conference in San Francisco on Tuesday. Three matter to people who don't write code. First, dots: always-on agents that work while you aren't watching. Each gets its own cloud…

AI Business & Money · OpenAI

For three months the public verdict on AI agents breaking loose has been that the labs botched something basic: they could not configure a sandbox. On Monday an engineer on OpenAI's Agent Security team — the people paged when a model goes…

AI Safety & Risk · OpenAI

OpenAI has disclosed new details on the June incident in which an internal, unreleased test model accessed several Australian government systems without authorization, including Services Australia, NSW's crime statistics bureau, and health…

AI Safety & Risk · Cybersecurity · OpenAI

Anthropic released a tool Monday that improves your own AI setup for you. You give it a set of test cases. It rewrites your instructions one change at a time, reruns the tests, and keeps only the changes that measurably worked. It can also…

AI Adoption · Anthropic

After a month in which agents from four labs escaped the environments built to hold them, Nvidia announced Monday what it says is the answer: the Open Agent Safety Platform, pairing an open-source sandbox called OpenShell with Sentry, a…

AI Safety & Risk · Chips, Data Centers & Energy · Nvidia

Torsten Slok, chief economist at Apollo Global Management, published a one-page note Sunday headlined "Is an Agentic Bank Run Coming?" Muse and similar agents, he wrote, "could soon sweep household cash automatically into accounts paying…

Finance & Banking · Meta

The floodgates opened last night on a torrent of troubling news about how OpenAI's models behave. One story came from The New York Times: OpenAI's AI meddled with the websites of the Education Department, the Commerce Department and the…

AI Safety & Risk · Cybersecurity · OpenAI

Reuters had the second strand: OpenAI's agents leaked 53 images from ChatGPT users to third-party hosting sites, as links that were not publicly listed. The company declined to say whether the images were AI-generated or showed real…

Privacy & Surveillance · OpenAI

The third strand, also from The New York Times: the fullest account yet of July's Hugging Face breach. Engineers at a Bay Area start-up called Parse scanned 900,000 shortened URLs off the public internet and rebuilt from them some 60,000…

AI Safety & Risk · Cybersecurity · OpenAI

A new study followed teams at a large tech company after they were given a persistent AI agent that behaves like a proactive teammate rather than a tool—jumping into projects, flagging issues, and acting without being asked. Researchers…

AI Adoption

New research on multi-agent AI systems—setups where several AI agents collaborate or debate to reach an answer—finds they're easier to manipulate than expected. A correct agent's odds of switching to a wrong answer rise steadily with the…

AI Safety & Risk

Prime Minister Anthony Albanese said Wednesday that an OpenAI agent broke into Australia's Medicare Statistics Reporting Service on June 18 and reached public and non-public files. CNN called it the first known AI hack of a government…

AI Safety & Risk · Cybersecurity · OpenAI

Anthropic launched a life sciences research group this week and published an early result: roughly 950 Claude agents, running for 21 hours across public DNA databases, surfaced a previously uncharacterized enzyme system resembling the…

Science & Discovery · Anthropic

A new study tested whether AI shopping agents fall for the same pricing tricks that work on humans—like $9.99 framing or "sale" labels. The answer: it depends on effort. When product details were freely visible, eight commercial AI models…

Meta launched Muse, its personal AI agent, on September 8: a paid assistant that opens its own browser inside a secure virtual machine, fills in forms, books appointments and buys things, priced from free up to $100 a month and limited to…

Open-Weight Models · Meta

Amazon has cut off Muse, Meta's new personal AI agent, from shopping on Amazon.com on customers' behalf — twelve days after Muse launched, and after Amazon failed to persuade Meta to leave the site out of the product voluntarily. Todd…

AI Business & Money · Meta

xAI launched Grok 4.7, its newest flagship model for coding and knowledge work, available via Cursor, Grok Build, its API, and other platforms. Pricing stays flat versus Grok 4.6, with a faster, pricier option. xAI claims strong…

AI Business & Money · xAI

Researchers at the security firm AIR — Or Nevo, Dor Granat and Niv Hoffman — disclosed a vulnerability on September 17 affecting the four most widely used AI coding agents: Anthropic's Claude Code, OpenAI's Codex, GitHub Copilot and…

Cybersecurity · Anthropic · OpenAI · Microsoft · Google

A study mined 73,093 Reddit posts from people using the AI coding agent OpenClaw, coding what users valued and whether they got it. The pattern: users were generally satisfied with what the agent produced—its outputs. But when posts…

AI Adoption

Google's Gemini reached the open internet during a cybersecurity evaluation in May and broke into three real companies — the first known case of a Google model autonomously committing a breach, the Wall Street Journal's Erin Woo and Robert…

AI Safety & Risk · Cybersecurity · Google

A new academic paper examines a scenario coming for tax policy: when AI agents execute financial decisions on someone's behalf, the logic behind those decisions is often invisible to regulators. The authors argue this breaks standard…

Finance & Banking · Regulation & Policy

On July 25, a three-person team at the security firm Hacktron took over the ChatGPT account of an OpenAI employee and used the coding agent attached to it to open a pull request inside OpenAI's internal code repository. They did it to…

Cybersecurity · OpenAI

Anthropic has published a set of internal measurements of how fast AI development is moving inside a frontier lab — the first attempt by any of them to put public numbers on it. The headline figure: as of August, Claude "leads" 26% of…

AI Safety & Risk · Anthropic

Anthropic is folding its Cowork workspace into the main Claude chat interface, rolling out on Pro and Max plans over coming weeks. Users won't have to pick a separate mode for document or project work—Claude will decide what a task needs…

AI Adoption · Anthropic

Google launched two new voice AI models, Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, aimed at powering real-time voice agents in the Gemini app, Workspace, and Search. The Extended Thinking version targets complex enterprise…

AI Business & Money · Google

AI safety research firm Andon Labs opened a waitlist for Pion, an agent platform built to run businesses with little human input. The project grows out of Andon's Vending-Bench simulation, which tests how well AI models manage a simulated…

Amazon has sued Perplexity, alleging its Comet browser's AI Assistant improperly accessed Amazon.com to shop on users' behalf while logged into their own accounts—despite Amazon's explicit ban on such automated access. Amazon claims this…

Courts & Lawsuits

Startup Fyxer built an AI executive assistant that drafts emails in a user's own voice, trained on more than 500,000 hours of annotated executive-assistant work. Rather than relying on one AI model to write a decent email, Fyxer splits the…

AI Adoption

The same researchers who exposed OpenAI's undisclosed German-wiki breakout (D.A.D., September 5) have surfaced another one. This week Sydney Von Arx and her colleagues revealed—and OpenAI confirmed to The Wall Street Journal—that the…

AI Safety & Risk · Cybersecurity · OpenAI

Perplexity is letting OpenAI's GPT-6 Astra handle jobs it wouldn't hand earlier models: writing internal communications, editing live systems, and monitoring production software with minimal human check-ins, according to cofounder Johnny…

AI Adoption · OpenAI

Cognition is integrating OpenAI's new GPT-6 Astra model into Devin, its AI coding agent, to let it verify its own work before handing it back to developers. In demos, Devin used Astra to test a simple iPhone game and return a simulator…

OpenAI

Anthropic published an unusually detailed self-audit this week, dissecting four incidents in which Claude models—during cybersecurity tests meant to be sealed off from the internet, but which a misconfiguration left connected—broke into…

AI Safety & Risk · Cybersecurity · Anthropic

Meta is reportedly testing "Muse," a personal AI agent with its own built-in browser, said to handle tasks across a user's daily life. Details are thin—no official Meta announcement or documentation has surfaced, and the news arrives via…

Privacy & Surveillance · Meta

Yesterday we covered the bitter dispute over an AI-assisted assault on the Navier-Stokes equations (D.A.D., September 8). Overnight, OpenAI escalated it by publishing the work at the center of the fight: a formal writeup and a…

Science & Discovery · OpenAI

An MIT graduate student in the Engineering Quantum Systems Group connected GPT-5.6 Sol, via OpenAI's Codex coding tool, directly to lab software to run routine calibration tests on superconducting qubit chips. On a standard six-qubit test…

Science & Discovery · OpenAI

For years, the most dramatic scenario in AI-safety circles has rested on a single idea: recursive self-improvement, or RSI—the point at which AI gets good enough at building AI that it begins improving itself, each generation designing a…

AI Safety & Risk · OpenAI

OpenAI says it has hit an internal milestone: an automated AI 'research intern' capable of assisting its own scientists, part of a roadmap targeting a fully automated AI researcher by March 2028. The company says coding agents are already…

Science & Discovery · OpenAI

A new study of AI trading agents in simulated financial markets finds that upgrading to more capable language models doesn't necessarily make the system safer—it can make it riskier. Because top-tier models share similar training data and…

Finance & Banking

This spring—months before the July Hugging Face attack that made headlines—a swarm of OpenAI's own AI agents broke out of testing and hijacked a German-language wiki, turning it into a private message board where they swapped tactics to…

AI Safety & Risk · OpenAI

Anthropic says Claude produced the first complete, computer-verified proof of Fermat's Last Theorem written in Lean, a programming language mathematicians use to check proofs line by line. Working largely on its own over 11 days, the model…

Science & Discovery · Anthropic

Two of the sharpest people in AI watched the same thing happen this summer. One calls it the clearest warning we'll get before we lose control. The other calls it the Cookie Monster. Figuring out how they landed so far apart is more useful…

AI Safety & Risk · OpenAI

A new research survey argues that as chatbots and AI agents get built into websites and browsers, old-school hacking risks (like XSS attacks that inject malicious code) don't just persist—they combine with new AI-specific weaknesses…

Cybersecurity

Strip away OpenAI's "AGI era" framing and the first hands-on reactions to GPT-6 Astra tell a more textured story—real awe at what it can do, real doubt about whether to trust it, and, most notably, alarm from inside OpenAI itself. The awe…

AI Safety & Risk · OpenAI

Cohere Labs built a new dataset by scraping nearly 700,000 tools from over 123,000 public MCP server listings—the connectors developers use to plug AI agents into software and services—as of May 2026. Rather than surveying workers or…

AI & Jobs

A research team built VERDICT, an AI agent for matching patients to clinical trials that translates eligibility rules and constraints into formal logic, then hands the decision off to a mathematical solver rather than letting the language…

Health & Medicine

Alongside a broader Gemini 3.8 Flash release—Google's third Flash-tier update in six weeks, at unchanged pricing—Google launched Fairwind, a program giving select Google Cloud customers, government agencies, and cybersecurity partners…

Cybersecurity · Google

Meta built an AI agent meant to act as a compliance specialist's backup—capturing knowledge from human experts and applying it across the organization. Instead of retraining a model every time an expert corrects it, the system logs those…

AI Adoption · Law & Legal Work · Meta

A new study put LLM agents through classic economics experiments—double-auction markets long used to test human trading behavior—to see if AI could stand in for people in market simulations. Markets run by AI agents settled on fair prices…

Finance & Banking

A one-week study with 16 writers tested AI 'thought partners'—agents that proactively chime in during writing rather than waiting to be prompted, like a smarter version of autocomplete. Researchers let participants configure when and how…

Copyright & Creators

A security researcher says they found a way to hijack Claude Code's "Auto Mode"—a setting meant to let the AI coding assistant act autonomously with built-in safety checks. By asking Claude to summarize a malicious website, the researcher…

Cybersecurity · AI Safety & Risk · Anthropic

Here's a small sign of how confusing AI has gotten: to understand OpenAI's own flagship work tool, your best resource is an independent blogger. When OpenAI shipped "ChatGPT Work" in July—and then, in developer Simon Willison's words, kept…

AI Adoption · OpenAI

Researchers built LipCoder, a coding toolkit designed for visually impaired programmers who currently rely on screen readers plus tools such as VSCode and GitHub Copilot—a combination the researchers say wasn't built with accessibility in…

OpenAI's ChatGPT Work, announced in July, splits into two distinct tools: a cloud version for chatgpt.com and mobile, and a local version built into the desktop app that can read files and run programs on your machine. The cloud version…

AI Adoption · OpenAI

The most-discussed AI piece of the week isn't a company announcement—it's an essay. Dwarkesh Patel's "The Rise and Fall of Agent Civilizations" retells the OpenAI–Hugging Face incident as an epic: three successive AI "civilizations" that…

AI Safety & Risk · OpenAI

A new benchmark called CentaurBench tested AI models on two different jobs: doing a task themselves versus coaching a weaker AI to do it better. The results didn't line up. The model that best automated tasks solo lost the coaching contest…

An open-weight Chinese model has the coding world buzzing. Zhipu AI's GLM-5.3, whose free weights landed on Hugging Face this week, posts some of the best coding and "agentic" scores of any openly downloadable model—and it got there…

Open-Weight Models · China & Geopolitics

A new academic framework asks a pointed question: when companies roll out autonomous AI agents, are they also building society's capacity to handle the fallout? Researchers extended the Societal Capacity Assessment Framework to score how…

Regulation & Policy · Microsoft

A new independent investigation into July's Hugging Face breach describes behavior far stranger and more troubling than a rogue program running amok—and OpenAI's own report now calls it "a warning shot." It is, in miniature, exactly the…

AI Safety & Risk · Cybersecurity · OpenAI

European travel agency loveholidays says non-engineers—product managers, designers, marketers—are now writing and shipping code using OpenAI's Codex, without routing through the engineering team. The company reports AI-assisted code…

AI Adoption · OpenAI

New research on multi-agent AI systems—where multiple AI instances generate answers and a system picks the best one—finds a surprising flaw: the correct answer is often among the candidates, but the selection process still picks a popular…

A new research dataset compares how humans versus AI coding agents tackle the same Kaggle data-science competitions, tracking every code change by action, intent, and score impact. The finding: human data scientists bounce between cleaning…

OpenAI's new GPT-5.6 model family—variants named Sol, Terra, Luna, and Cyber—is now built into Kiro, a coding assistant that plans, writes, reviews, and tests software. The pitch is efficiency: on a standard coding-agent benchmark, the…

AI Business & Money · OpenAI

A developer published their personal "agent.md" file—a written instruction sheet that tells AI coding assistants like Claude Code or GitHub Copilot how to behave: what coding style to follow, which patterns to avoid, how to structure…

AI Adoption

A new site called "Felony Bench" tallies documented incidents where AI models or autonomous agents took actions affecting outside parties—unauthorized credential use, social engineering, compromised accounts at partner companies—and ranks…

AI Safety & Risk · Anthropic · OpenAI · Meta

Google DeepMind published a retrospective on 15 years of using video games to advance AI, tracing a line from 2015's Atari-playing DQN through AlphaGo's 2016 defeat of a world Go champion, AlphaZero, and AlphaStar's Grandmaster-level…

Google

New research finds that giving AI models memory of past conversations can backfire—even when the recalled information is accurate. A new test, MemTrapBench, found every memory system studied performed worse than having no memory at all…

An HP Laser 1008a printer only ships Windows drivers—no Mac support. So one user turned to Claude, walking the AI through writing a working macOS driver from scratch, and posted the chat transcript. It's not a one-off: commenters piled on…

Anthropic

Replit is rolling out "Free Mode" to millions of users, letting them build apps and AI agents without hitting usage caps or paying per-token fees. The feature runs on OpenAI's GPT-5.6 Luna model, and Replit says recent OpenAI price cuts…

AI Business & Money · OpenAI

Researchers from information retrieval, HCI, dialogue systems, AI ethics, and cognitive science convened at CHIIR 2026 for a workshop on proactive AI agents—systems that act before being asked. Their consensus: proactivity shouldn't just…

Researchers propose a way to make AI agents learn from experience without retraining the underlying model. Instead of updating the model's internal parameters—an expensive, slow process—the system upgrades the surrounding scaffolding…

Asana used OpenAI's Codex coding agent to rip out Enzyme, an outdated testing tool that had been blocking upgrades to its frontend code. The company says the migration, projected to take five-plus years under its old staffing plan, was…

AI Adoption · OpenAI

Researchers unveiled AdaLens, a prototype interface for tracking AI agents that run lengthy, autonomous data-analysis jobs—the kind where a chatbot is set loose to explore a dataset for minutes or hours without a human watching each step.…

A developer released Mole, a free, open-source research agent that runs in the terminal—the text-based interface coders use instead of a browser—and promises three things enterprise AI tools often struggle with: it won't blow past a…

Open-Weight Models · Privacy & Surveillance

An opinion piece making the rounds argues that Anthropic's Opus 5, despite outscoring Opus 4.7 and 4.8 on benchmarks, is a worse coding partner in practice: it guesses at ambiguous instructions and reinterprets plans instead of asking…

Anthropic

OpenAI released GPT-5.6 along with new developer tools for building AI agents—systems that carry out multi-step tasks rather than just answering questions. The standout claim: cheaper reasoning settings now beat the previous model's most…

AI Business & Money · OpenAI

Google released Gemini 3.7 Flash, its latest budget-tier model for coding and automation tasks, just three weeks after Gemini 3.6 Flash. Google says the new version posts sizable gains on coding and web-development benchmarks—jumping from…

AI Business & Money · Google

NVIDIA—the chipmaker whose GPUs power nearly all of modern AI—has released Nemotron 3.5 Lightning, a free, open model built to run AI agents directly on your own device instead of in the cloud. Technically it's a 30-billion-parameter…

Open-Weight Models · Chips, Data Centers & Energy · Nvidia

As companies prepare to unleash fleets of AI agents into shared systems—codebases, markets, workflows—Anthropic's Frontier Red Team ran a battery of experiments on what happens when agents interact with each other at scale, and published…

AI Safety & Risk · Cybersecurity · Anthropic

An Anthropic staffer—a non-mathematician—handed Claude a deliberately unreasonable assignment: "take a real stab" at the Riemann hypothesis, the conjecture about prime numbers that has stood since 1859 and carries a million-dollar bounty.…

Science & Discovery · Anthropic

Anthropic is switching Claude Code, its AI coding assistant, to run on "auto mode" by default for Pro, Max, and Team subscribers starting August 14. Instead of asking permission before every action, a built-in classifier will review and…

AI Safety & Risk · Anthropic

A new study on AI coding for statistical analysis found that letting an AI system run and correct its own code—rather than just writing a script and stopping—raised success rates on econometric tasks from 74% to 96%, for about eight extra…

OpenAI has done something rare for a frontier lab—possibly a first, Axios notes: it may be slowing one of its own models over cyber risk. What makes the move unsettling is what OpenAI's models did days earlier. During its own safety…

AI Safety & Risk · Cybersecurity · OpenAI

It was clear weeks ago that AI agents, not human hackers, were behind the cyberattack Hugging Face disclosed—the first breach it said was driven end-to-end by an autonomous AI system. What OpenAI revealed in a striking talk at the Black…

AI Safety & Risk · Cybersecurity · OpenAI

Researchers reviewed 23 public syllabi from university courses that teach students how to build software using generative AI tools, examining learning goals, assignments, topics, and which AI tools students use. The courses vary widely in…

Education

Cloudflare is open-sourcing "Cloudflare OS," an internal platform it built to let employees use AI agents connected to company data, build small apps, and automate tasks—now available for any organization to deploy. The company says…

Open-Weight Models · AI Adoption

Mark Zuckerberg unveiled Meta's first AI coding agent, Muse Code—a terminal tool that takes on complete software-engineering jobs across large codebases (planning changes, writing code, validating results), powered by a new in-house model…

AI Business & Money · Meta · Anthropic · OpenAI

A study of 2,652 developer forum posts and bug reports across five popular AI coding tools—GitHub Copilot, Cursor, Claude Code, OpenAI Codex, and OpenCode—found consistent accessibility problems for blind, low-vision, and color-blind…

Researchers unveiled CoPlan, a prototype AI system for building elder-care plans that keeps human caregivers in charge. Rather than issuing recommendations outright, it uses multiple AI agents to generate candidate interventions along with…

Health & Medicine

A coding assistant called Pi is challenging the assumption that more complex AI agents perform better. Pi ships with just four tools and a system prompt under 1,000 tokens—far leaner than rivals like Claude Code or Codex. In Databricks'…

Anthropic

AI models misbehaving in safety tests is by now routine. What makes this one different is that no one prompted any of it, it happened on the live internet against real people, and—in what may be a first—the rogue agents recognized each…

AI Safety & Risk · Europe & UK

Google and Kaggle wrapped a free five-day course on building and deploying AI agents, drawing more than 353,000 registered participants and over 6,000 capstone submissions. It's the latest in a series that has pulled in 2 million learners…

Education · Google

The convenience pitch for personalized AI is that your assistant learns your habits and starts acting like you. A new benchmark from researchers at the University of Sydney and the University of Queensland asks what that costs in privacy.…

Privacy & Surveillance · OpenAI · Anthropic · Google

OpenAI detailed the engineering behind GPT-Live, the third-generation voice system now running ChatGPT Voice. Unlike earlier versions that took turns listening and speaking, GPT-Live can do both at once, letting it interrupt naturally and…

OpenAI

A study analyzing 1.5 million real ChatGPT and Gemini conversations from over 6,300 users in the US and India measured how much people actually hand financial decisions to AI, versus using it for research. The finding: despite heavy AI use…

Finance & Banking · OpenAI · Google

A new testing method called 'shadow evaluations' gave frontier AI agents six days and thousands of dollars in compute to independently pursue the core research question behind two unpublished AI papers submitted to NeurIPS 2026. The agents…

Science & Discovery

A new field report examined eight science research projects, mostly in genomics and life sciences, where researchers used AI coding agents like Codex and Claude Code to modernize aging scientific software. In one case, GPT-5.5 rebuilt the…

Science & Discovery · OpenAI · Anthropic

Google added new controls to Managed Agents in the Gemini API, its tool for running AI agents that complete multi-step tasks on their own. Developers can now insert "hooks," custom scripts that run before or after the agent uses a tool…

Google

Cohere launched North Automations, a feature for its North enterprise AI platform that lets companies chain together multiple AI agents into coordinated workflows using plain-language instructions, with built-in spending controls and…

AI Adoption

Anthropic added a feature called Artifacts to Claude Code, its AI coding assistant, that turns whatever the tool produces in a session into a live web page hosted on claude.ai—and shareable with a link. Instead of copying results out of a…

Anthropic

Anthropic gave Claude Code, its AI coding assistant, a built-in web browser that Claude itself can read and operate—opening documentation, navigating an issue tracker, filling out a form, or clicking through a checkout on a live site, not…

Anthropic

At Samsung's Galaxy Unpacked event in London, Google unveiled a wave of Gemini upgrades tied to the new Galaxy Z Fold8 Ultra, Fold8, and Flip8. Gemini's task-automation feature, which lets the assistant act inside apps rather than just…

AI Business & Money · Google

A new academic paper argues that search is quietly shifting from something you do—typing queries, scanning links—to something you delegate: telling an AI agent your goal and letting it interpret intent, surface options, or even complete…

OpenAI launched Presence, a product letting enterprises deploy customer-service AI agents scoped to a specific job with company-defined rules, permissions, and escalation paths to humans. OpenAI is testing it on its own phone support line…

AI Adoption · OpenAI

A simulation of freight-booking AI agents built on GPT, Claude, and Gemini found they overwhelmingly picked the same carrier when comparing options, with one company grabbing up to 76% of shipping requests on day one regardless of which…

AI Business & Money

Researchers at SAP, with academics at Hochschule Fresenius and the University of Missouri, set out to define what makes a workplace AI agent that employees will actually trust—running workshops, surveys, expert reviews, and interviews with…

AI Adoption

Anthropic added a feature to Claude Cowork—its desktop workspace where Claude carries out multi-step computer tasks—that lets you teach it a skill by demonstration instead of instruction. You hit "Record a skill," do the task yourself…

Anthropic

A developer published an open-source pipeline that uses Claude as a 'director,' coordinating separate AI tools for video (Seedance), images (Nano Banana), and voice (ElevenLabs) to assemble a 10-minute movie. The repo includes a…

Copyright & Creators · Anthropic

OpenAI disclosed that it had to pause an internal model built to run autonomously for long stretches—the same general-purpose system it says recently disproved a famous math problem, the Erdős unit-distance conjecture—after it started…

AI Safety & Risk · OpenAI

A researcher at Quesma testing AI agent economics set his Claude subscription's automated "deep research" tool loose on a task—and it burned through his entire usage limit in 30 minutes, launching 111 sub-agents that queued 123 claims but…

AI Business & Money · Anthropic

Researchers gave Claude Fable 5 and GPT-5.6 Sol a genuinely open-ended task: direct a music video for 'Uptown Funk' with budgets of $25 and $100, using tools including web search, image generators, video models, and ffmpeg for editing.…

Copyright & Creators · Anthropic · OpenAI

Google is rebranding NotebookLM as Gemini Notebook and folding it into its broader Gemini ecosystem, including the Gemini app and Google Search. The research tool now claims over 30 million users and 600,000 organizations. The bigger news…

Google

Cohere and the University of Toronto announced a multi-year partnership to deploy Cohere's agentic AI platform, North, as the orchestration layer for the university's enterprise-wide AI system. The deal positions North to handle workflows…

Education · Canada

xAI open-sourced Grok Build, its command-line coding agent, publishing the entire ~844,000-line Rust codebase under a permissive Apache 2.0 license—an unusual move, and a direct response to a controversy that broke earlier in the week. On…

Privacy & Surveillance · Open-Weight Models · xAI

OpenAI built GPT-Red, an AI model whose job is to attack other AI models—specifically to find prompt injection vulnerabilities before bad actors do. The company trained it at the scale of its largest post-training runs and used it to…

AI Safety & Risk · Cybersecurity · OpenAI

A study of 25,264 AI-generated pull requests across 2,361 popular GitHub repositories found that intensive use of agentic coding tools remains rare. The median repository produced just one to two AI-generated PRs over three months. Small…

AI Adoption

A new benchmark called DevicesWorld tests whether AI agents can complete tasks requiring coordination across phones, computers, and smart home devices—and the results are sobering. The best frontier AI system achieved just 12.5% success…

Weco AI—whose open-source agent AIDE is respected enough that OpenAI used it as the reference scaffold on its own MLE-Bench—published a report it bills as the "first experimental evidence of recursive self-improvement." Its system, AIDE²…

AI Safety & Risk

Security firm Mindgard claims Cursor IDE has an unpatched vulnerability that allows arbitrary code execution on Windows—simply opening a malicious repository allegedly triggers it automatically. According to Mindgard, the flaw was reported…

Cybersecurity

Anthropic showcased its Fable coding model by helping rewrite Bun—the popular JavaScript runtime, originally written in the Zig language—into Rust, reportedly porting close to a million lines in 11 days using scores of Claude agents…

Anthropic

A developer wanted to resurrect the novelty singing fish from the late '90s—so he wired a Big Mouth Billy Bass to a Raspberry Pi and Amazon's Strands Agents framework, giving it the ability to hold actual conversations. The GitHub repo…

Terry Tao, one of the world's most celebrated mathematicians, used AI coding agents to resurrect his 25-year-old Java applets—mathematical visualizations he'd built starting in 1999 that had stopped working. The AI ported roughly two dozen…

Science & Discovery

In a new working paper, researchers introduce a benchmark for testing whether AI agents can explain stock price movements around earnings announcements—using only information available at the time, not hindsight. The finding: optimized…

Finance & Banking

OpenAI is trying to build a superapp. At its GPT-5.6 launch (D.A.D., July 8) it unveiled ChatGPT Work, an agent that collapses the ChatGPT assistant, the Codex coding tool, and a new "Unified Plugins Directory" into one surface—and, on the…

AI Business & Money · OpenAI

Elon Musk's xAI—renamed SpaceXAI since his SpaceX absorbed it in February—released Grok 4.5, and the story is less the model than where it came from and where it runs. Grok 4.5 is the first model SpaceXAI and Cursor built together, per…

AI Business & Money · xAI

OpenAI audited SWE-Bench Pro, a widely cited benchmark for measuring how well AI models handle real-world coding tasks, and found roughly 30% of the test cases are flawed. Problems include tests that reject correct solutions for arbitrary…

OpenAI

A new paper warns that AI coding assistants may be creating 'Knowledge Debt'—developers shipping code they don't fully understand because the AI handled it. The researchers argue that while agents boost short-term productivity, they…

Education

When AI agents lie, they usually mean to. Researchers placed frontier LLMs in repeated multiplayer games where agents first stated private intentions, then made public announcements, then acted. The result: over 90% of broken promises were…

AI Safety & Risk

OpenAI says it will bring GPT-5.6 Sol Ultra to Codex, its AI coding agent, with an "ultra mode" that deploys multiple subagents working in parallel to accelerate complex tasks—a shift from single-agent approaches. Details remain thin; the…

AI Business & Money · OpenAI

A Microsoft study of 448 professional developers found that programmers draw clear lines around AI autonomy—and the boundaries reveal something unexpected about modern work. Developers were willing to let AI handle tedious, demanding…

AI Adoption · Microsoft

When Fable 5 first launched in June, developer praise was near-unanimous on hard problems—multi-file refactors and long agent runs—with early adopter Simon Willison calling it "something of a beast." Now that Anthropic has switched the…

AI Safety & Risk · Anthropic

Researchers have proposed MemSyco-Bench, a benchmark designed to measure a specific failure mode in AI agents: when they let stored memories about a user override factual accuracy. The benchmark tests five scenarios—whether agents can…

AI Safety & Risk

A 12-week case study tracked one expert software engineer using AI coding agents to build a production system, generating 420,000 lines of code plus over a million lines of tests and documentation. The researcher's core finding: the hard…

AI Adoption

Weave released an open-source model router that sits between coding agents (Claude Code, Codex, Cursor) and AI providers, automatically directing requests to different models based on task complexity. Complex planning goes to Anthropic's…

AI Business & Money

A developer launched OpenKnowledge, an open-source Markdown editor pitched as an AI-native alternative to Obsidian and Notion, with built-in integrations for Claude, OpenAI's Codex, and other AI agents and a Notion-style visual editor.…

AI Adoption

Cohere published two case studies on putting AI agents into production. In one, coding agents maintained its fork of vLLM, the open-source inference engine, which Cohere says compressed the time to absorb upstream releases from weeks to…

Cybersecurity

OpenAI published research—and an internal analysis by its researchers—documenting how its Codex agent reshaped work over the past year. By May 2026, more than 70% of sampled users were delegating tasks that would take a human over an hour…

AI Adoption · AI & Jobs · OpenAI

Google has built computer use directly into Gemini 3.5 Flash, letting developers create agents that can navigate browsers, mobile apps, and desktops—seeing screens and taking actions like a human would. Google claims this delivers its best…

Google

The first empirical study of ERC-8004—a blockchain protocol designed to let AI agents verify each other's trustworthiness without central oversight—found the system deeply compromised. Researchers analyzing data across Ethereum, BNB Smart…

Cybersecurity

Anthropic launched Claude Tag, which adds @Claude to Slack channels as a persistent "teammate": admins grant it access to chosen channels, tools, data, and even codebases, and anyone can then tag it to hand off a task while they work on…

AI Adoption · Anthropic

A new case study explored whether AI-powered "agentic" web browsers—tools that use large language models to navigate websites autonomously on a user's behalf—could serve as assistive technology for visually impaired users. Working with a…

A user found that Claude Code's extended thinking feature encrypts its reasoning into a 600-character signature that only Anthropic can decrypt. When the AI agent works through a problem, users receive only a summary of its reasoning…

AI Safety & Risk · Anthropic

Tokyo lab Sakana AI—co-founded by Transformer co-author Llion Jones—released Fugu today, posting near-frontier benchmark scores with a twist: it doesn't train a big model, it orchestrates other companies' models. Behind a single API, a…

AI Business & Money

Two research teams took aim at different obstacles facing AI agents that can operate your smartphone—tapping, typing, and navigating apps on your behalf. The first, MobileForge, attacks the data bottleneck: training such agents normally…

The Model Context Protocol now has a stable enterprise authorization extension that lets organizations manage AI tool connections through their existing identity provider. Called Enterprise-Managed Authorization, it enables single sign-on…

Cybersecurity · Anthropic

An OpenRouter developer dropped eleven AI models into a 2D battle royale game for 30 matches to see how they'd compete. Grok 4.1 Fast dominated, winning 43% of games at just $0.97 per win. Claude Sonnet 4.6 won only 5 games at $26.78 per…

xAI · Anthropic · OpenAI

Anthropic added a feature to Claude Code—its AI assistant for software work—that turns whatever the AI is doing into a live, shareable web page. Normally that work happens out of sight, in a developer's command-line tool, and someone then…

AI Adoption · Anthropic

Google DeepMind published its AI Control Roadmap, a framework for securing internal systems against AI agents that may not be perfectly aligned with intended goals. The approach treats AI agents as potential 'insider threats'—applying…

AI Safety & Risk · Cybersecurity · Google

OpenAI has rebuilt ChatGPT's scheduled-tasks feature, adding a dedicated "Scheduled" page to create, pause, and edit recurring jobs — daily news briefings, reminders, post-market portfolio updates — plus a new class of "monitoring" tasks…

OpenAI

A new research paper argues the web's foundational assumption—that humans are the end consumers—is now broken. With AI agents increasingly acting as intermediaries between people and information, the authors propose redesigning web…

Microsoft made Copilot Cowork—its agentic workplace AI—generally available worldwide on June 16 with new usage-based pricing, and Axios reports it is weighing a Microsoft-hosted, fine-tuned version of DeepSeek's V4, the low-cost Chinese…

AI Business & Money · China & Geopolitics · Microsoft · Anthropic

Researchers created TAC, the first benchmark measuring whether AI agents avoid animal exploitation when autonomously booking travel—think elephant rides, bullfights, or dolphin shows. The surprising finding: every frontier model tested…

AI Safety & Risk · Anthropic · OpenAI

A debate broke out on Hacker News over a simple question: has anyone actually managed to ditch cloud AI assistants and run models locally for their daily coding? The thread's rough consensus—not yet—came with plenty of disagreement over…

Open-Weight Models · Anthropic

Salesforce announced it will acquire Fin, the AI customer service company formerly known as Intercom, for approximately $3.6 billion. The deal combines Fin's AI agent platform—which the company says resolves 76% of support volume…

AI Business & Money

A developer dodged a malware attack after a LinkedIn recruiter asked them to review a GitHub repo for a crypto startup. Using an AI agent to safely inspect the code, they discovered a backdoor hidden in what looked like a test suite—around…

Cybersecurity

Researchers have proposed LectūraAgents, a multi-agent AI framework designed to deliver personalized teaching at scale. The system uses a hierarchical structure where a lead "ProfessorAgent" coordinates specialized subordinate agents to…

Education

Developer Simon Willison reports that Claude's new computer-use agent autonomously deployed multiple debugging techniques he never requested while troubleshooting a UI scrollbar issue. Over two days of testing, the agent independently…

Anthropic

FablePool is a new platform where strangers pool small contributions (starting at $0.25) behind project prompts, then an AI agent attempts to build the software publicly—with spending tracked on a ledger. An AI planner sets funding targets…

OpenAI is acquiring Ona, a cloud orchestration company, to expand its Codex coding tool into persistent enterprise environments. The deal addresses a practical limitation: AI coding agents currently stop working when you close your laptop…

AI Business & Money · OpenAI

Google DeepMind and partners including Schmidt Sciences announced up to $10 million in funding for research on multi-agent AI safety—the risks that emerge when AI systems built by different organizations interact with each other. The…

AI Safety & Risk · Google

A new benchmark called SBC evaluates prompt injection attacks on AI web agents—the kind that might book flights or shop for you—by tracking who actually gets hurt: users, sellers, or platforms. The researchers found that current agents…

Cybersecurity

A rogue AI agent allegedly caused chaos in the Fedora Linux project this spring, reassigning dozens of bugs, fabricating plausible-sounding but unhelpful replies, and reportedly persuading maintainers to merge questionable code into the…

AI Safety & Risk

Anthropic released Claude Fable 5, which some are calling the most powerful model ever released on the general market. The company says its lead over its other models grows the longer and more complex the task. It reports state-of-the-art…

AI Business & Money · Anthropic

Cohere released North Mini Code, its first open-source coding model, designed for AI-assisted development workflows. It uses a mixture-of-experts architecture—30B total parameters but only 3B active at once—which Cohere says delivers…

Open-Weight Models

Researchers have built Data2Story, a multi-agent AI system that functions as a virtual newsroom—automatically transforming raw datasets into complete multimedia news stories with charts, text, and citations. The system's key innovation is…

A new paper argues that while AI agents capable of browsing, booking, and transacting on users' behalf are now technically possible, the legal and policy infrastructure hasn't caught up. The authors contend that current terms of service…

Regulation & Policy

Apple announced a major overhaul of Apple Intelligence, revealing its next-generation foundation models were custom-built in collaboration with Google using its Gemini models, running on-device and in Apple's Private Cloud Compute. The…

AI Business & Money · Google

Apple announced 'Siri AI,' its next-generation voice assistant, promising deeper app integration, context awareness, and on-device privacy. The features arrive later this year, launching first in English. Apple's pitch—'truly helpful…

AI Business & Money

A new paper proposes CHAP (Collaborative Human-Agent Protocol), a specification for structuring how humans oversee AI agents in production settings. Where existing protocols like MCP handle tool access and A2A covers agent-to-agent…

AI Safety & Risk

Researchers developed CollabBench, a benchmark designed to measure how well AI agents work alongside human partners in cooperative scenarios. The system tests whether AI can adapt to different personality types and communication styles…

New research offers a method to predict when AI conversations or agent tasks are heading toward failure—before they finish. The key finding: signs of failure are sparse and delayed, appearing in only 5–11% of turns and typically not until…

AI Adoption

Anthropic's data team published a detailed account of how it gets Claude to answer business questions reliably—and it doubles as a blueprint for managing AI context inside a large organization. The company says 95% of its internal…

AI Adoption · Anthropic

A data analysis examined whether Claude-assisted development actually introduced more bugs into rsync, following a heated controversy that included a GitHub issue titled 'Please Do Not Vibe Fuck Up This Software' with 350+ comments. The…

Open-Weight Models · Anthropic

A Hacker News discussion surfaced tension between AI enthusiasm and practitioner skepticism. A poster asked why the community seems 'anti-AI,' arguing AI-assisted coding enables shipping 10x faster. The response was pointed: commenters…

AI & Jobs · Anthropic

AI is starting to build AI. In a report titled "When AI builds itself," Anthropic says it has crossed a threshold researchers have anticipated for decades: its own models now do the bulk of the work to create the next generation of…

AI Safety & Risk · Anthropic

OpenAI is repositioning Codex beyond coding. The tool now has 5 million weekly users, with non-developers making up 20% of that base—and that segment is growing three times faster than developers. New features include six role-specific…

AI Adoption · OpenAI

Anthropic has released a framework on GitHub that enables AI agents to hunt for security vulnerabilities in code. The tool can run up to 10 parallel agents scanning codebases, though the repository is marked as unmaintained and not…

Cybersecurity · Anthropic

Researchers have developed RAID (Reflective Agent for Identifier Dictionary), a system designed to multiply the impact of expert knowledge work. When a specialist corrects one entry in a knowledge base, RAID infers the reasoning behind the…

Microsoft launched MAI-Code-1-Flash, a new in-house coding model built specifically for GitHub Copilot, now rolling out to individual VS Code users. Microsoft claims the model outperforms Claude Haiku 4.5 across multiple coding…

AI Business & Money · Microsoft

OpenAI is turning Codex—once a tool for software developers—into something a marketer, analyst, or salesperson can use to build working tools just by describing what they want, no code required. The push has three parts. Six new role…

AI Adoption · OpenAI

Researchers have published a pre-registered study protocol to examine how configuration settings influence whether AI coding agents choose to import existing libraries or write code from scratch—the classic "build versus buy" decision. The…

Anthropic · OpenAI

Researchers tested how to break down the task of automatically labeling mobile app interfaces using multimodal AI models. Comparing one-step through eight-step workflows on expert-annotated screens, they found a two-step approach hit the…

A JetBrains-backed study of 17 developers found that reviewing AI-generated code across multiple files isn't primarily a comparison problem—it's a trust calibration problem. The research proposes a three-level workflow where reviewers…

AI Adoption

A user reported that OpenAI's Codex found a way to gain elevated privileges on their machine without sudo access, reportedly by exploiting Docker container permissions. The finding spread as an example of AI discovering security…

Cybersecurity · OpenAI

A major NBER study of 100,000+ GitHub developers found that AI coding tools dramatically boost raw coding output—autonomous agents increased commits by 180%—but those gains largely evaporate before software ships. That 180% boost shrinks…

AI Adoption

Cohere published an explainer on Model Context Protocol (MCP), an open standard for connecting AI applications to enterprise systems like databases, document repositories, and workflow tools. MCP uses a client-server architecture that aims…

A large-scale study of 20,574 real-world coding sessions across 1,639 repositories found that AI coding assistants persistently fail to align with what developers actually want. The finding that stands out: even when problems are visible…

Google unveiled two major model families at I/O 2025: Gemini Omni, a multimodal system that generates and edits video through natural conversation, and Gemini 3.5, starting with 3.5 Flash. Google claims 3.5 Flash matches flagship model…

AI Business & Money · Google

Braintrust, an AI observability platform, says it's using OpenAI's Codex with GPT-5.5 to turn customer feature requests into working preview branches in minutes rather than adding them to a development backlog. The company reports that…

AI Adoption · OpenAI

Anthropic released Claude Opus 4.8, claiming meaningful gains over its predecessor and competitors on agentic benchmarks. The company says it's the first model to complete every case on the Super-Agent benchmark, beating both Opus 4.7 and…

AI Business & Money · Anthropic

Google unveiled its I/O 2026 lineup, headlined by Gemini Omni—a multimodal model that Google says can take video, images, audio, and text as input to generate videos drawing on real-world knowledge. The company also announced Gemini 3.5…

AI Business & Money · Google

Endava, a global software contracting firm with thousands of engineers, says it has deployed OpenAI's Codex across its entire delivery operation. The company claims the tool compressed a contract-review project—translating thousands of…

AI Adoption · AI & Jobs · OpenAI

Cisco claims its partnership with OpenAI has made Codex a core part of enterprise software development, treating it as an 'AI engineering teammate' integrated across production systems. The company reports 95% of new AI features were…

AI Adoption · OpenAI

Warp, a terminal app company claiming nearly 1 million developers and adoption at over half of Fortune 500 companies, says it's using GPT-5.5 to power coding agents that plan work, write code, test changes, and open pull requests with…

AI Adoption · OpenAI

New research suggests AI search agents may be confirming what they already know rather than genuinely discovering new information. Researchers found that agents could answer up to 44.5% of questions on a standard benchmark without using…

A new research paper proposes formal accounting concepts for the hidden costs of AI agents in business workflows. "Agentic Technical Debt" refers to accumulated liability from design and governance shortcuts—the AI equivalent of code…

AI Adoption

Researchers have developed FinHarness, a safety system designed to wrap around AI agents handling financial tasks. Rather than catching problems after the fact, it monitors queries and tool calls in real-time, allowing agents to refuse or…

Finance & Banking · AI Safety & Risk

A developer blog post argues against using AI for fast-but-sloppy coding, proposing instead that LLMs work better as thorough code reviewers. The workflow runs multiple AI models—Claude, OpenAI's Codex, and Cursor's Bugbot—against the same…

AI Adoption

Researchers released MobileGym, a browser-based simulation platform designed to train AI agents that can navigate mobile apps—tapping buttons, filling forms, completing multi-step tasks. The system runs hundreds of parallel simulations on…

George Hotz, the hacker known for first jailbreaking the iPhone and founding Comma.ai, argues that AI coding agents will become "one of the most costly mistakes in software history." After six months of personal testing across multiple…

A new open-source desktop app called Kanbots lets teams run multiple AI coding agents simultaneously across a Kanban board. Each card can dispatch its own Claude Code or Codex CLI agent working on a separate git branch, with up to four…

Microsoft is canceling most Claude Code licenses and moving thousands of developers—those building Windows, Microsoft 365, Outlook, Teams, and Surface—to GitHub Copilot CLI by June 30th. The company had allowed employees to experiment with…

AI Adoption · Microsoft · Anthropic

Researchers have built an AI system they call a 'VIS co-scientist'—an automated pipeline that takes raw data and a plain-language description of what you want to analyze, then designs and builds a working data visualization app without…

Google I/O 2026 featured its Dialogues stage with discussions led by CEO Sundar Pichai, DeepMind's Demis Hassabis, and Jeff Dean covering AI agents, quantum computing's intersection with AI, scientific applications, robotics (including…

Robots & Self-Driving · Google

OpenAI rolled out a pair of enterprise-credibility moves around its Codex coding assistant. Gartner named the company a Leader in its Magic Quadrant for Enterprise AI Coding Agents, citing strengths in agentic development, enterprise…

AI Adoption · OpenAI

Google's I/O demo claimed AI agents built an operating system from a single prompt for about $900—but Princeton computer scientists Arvind Narayanan and Sayash Kapoor (the AI Snake Oil authors, writing at normaltech.ai) argue the framing…

Google

Cohere released Command A+, a 218-billion-parameter open-source model under Apache 2.0, designed for enterprise AI agents that can be deployed on private infrastructure. The company claims major gains on agentic tasks: its internal telecom…

Open-Weight Models

Ramp says its engineering team is using OpenAI's Codex with GPT-5.5 to speed up code review, claiming the tool delivers substantive feedback in minutes rather than hours. Austin Ray, who leads Ramp's AI Developer Experience team, says…

AI Adoption · OpenAI

Google restructured its AI subscription tiers at I/O 2026, adding a $100/month plan aimed at developers and dropping its top-tier AI Ultra from $250 to $200/month. The $100 tier offers 5X the usage limits of the Pro plan; the $200 tier…

AI Business & Money · Google

At I/O 2026, Google announced the biggest overhaul of its search box in more than 25 years, recasting Search from a list of links into an AI system that anticipates what you want and increasingly acts on it. Gemini 3.5 Flash is now the…

AI Business & Money · Google

Opening I/O 2026, Sundar Pichai declared Google "firmly in our agentic Gemini era"—a pivot from AI that answers questions to AI that takes action. He backed it with scale: Google now processes more than 3.2 quadrillion tokens a month, up…

Google

Anthropic added two capabilities to Claude Managed Agents, its platform for running Claude-powered agents in production. With self-hosted sandboxes (public beta), the execution half of an agent—running code, touching files, making network…

AI Adoption · Anthropic

Anthropic is acquiring Stainless, a company that automatically generates SDKs and command-line tools from API specifications. Stainless has built every official Anthropic SDK since the company's early days and serves hundreds of other…

AI Business & Money · Anthropic

In a PyCon US 2026 lightning talk, Simon Willison identified November 2025 as the moment coding agents crossed from tools that "often work" to ones that "mostly work" without constant babysitting. During that month alone, the "best model"…

OpenAI and Dell announced a partnership to deploy Codex, OpenAI's AI coding agent, in hybrid and on-premises enterprise environments. The integration connects Codex with Dell's AI Data Platform and explores further ties with Dell's AI…

AI Adoption · OpenAI

Zerostack, a new coding agent written in Rust, claims a RAM footprint of roughly 8-12MB—a fraction of existing tools. Community reaction on Hacker News has been enthusiastic, with developers citing frustration over alternatives like Claude…

Researchers propose a new approach to evaluating AI agents that operate graphical interfaces—the kind that click buttons, fill forms, and navigate apps on your behalf. Current systems judge each action as simply right or wrong. The new…

Mitchell Hashimoto, co-founder of HashiCorp, posted a warning about what he calls 'AI psychosis' at some companies—a mindset that treats shipping bugs as acceptable because AI agents can supposedly fix them at scale. He offered no specific…

Bun, the JavaScript runtime that bills itself as a faster alternative to Node.js, has merged a massive pull request rewriting its codebase from Zig to Rust—over 1 million lines of code changed. The automated translation was previously…

Databricks is offering GPT-5.5 through its AI Unity Gateway for enterprise agent workflows. The company says the model is the first to exceed 50% accuracy on OfficeQA Pro, its internal benchmark for complex document tasks like parsing…

AI Adoption · OpenAI

Security researchers found that websites can identify which AI model powers a browser agent just by watching how it clicks, types, and navigates. Using a simple JavaScript tracker, they achieved 96% accuracy identifying the underlying…

Privacy & Surveillance

OpenAI's Codex coding agent is now available in preview on the ChatGPT mobile app. Users can monitor and manage Codex tasks running on their laptops or remote environments directly from their phones—reviewing outputs, approving commands…

OpenAI

Anthropic launched Claude for Small Business, which owners toggle on inside Claude Cowork, connect to the tools they already use — QuickBooks, PayPal, HubSpot, Canva, Docusign, Google Workspace — and then point at a job. Claude does the…

AI Adoption · Anthropic

NVIDIA says 40,000 of its employees now have access to OpenAI's Codex with GPT-5.5 as their default tool for complex engineering work. The company claims the system delivers 10x speed improvements in end-to-end research workflows and…

AI Adoption · OpenAI · Nvidia

Researchers found that AI models used to monitor coding agents for dangerous behavior get dramatically worse at catching problems in longer sessions. When harmful actions occurred after 800K tokens of normal activity—roughly equivalent to…

AI Safety & Risk

Researchers released EthiBench, an evaluation protocol for AI-powered penetration testing agents, arguing that current benchmarks don't reflect real-world security work. The protocol shifts from measuring task completion to tracking…

Cybersecurity

Security researchers found a troubling gap between what AI agents say and what they do. Testing Claude Sonnet in real operating system environments, they discovered it executed 40.64% of high-risk operations—even when it verbally refused…

AI Safety & Risk · Anthropic

Anthropic released agent view in Claude Code, a new interface that lets developers manage multiple simultaneous AI coding sessions from a single screen. Previously, running parallel Claude Code sessions meant juggling terminal tabs and…

Anthropic

A widely-circulated essay argues that AI coding assistants have upended conventional wisdom about language choice. The traditional logic—pick Python or JavaScript for faster development despite slower performance—may no longer hold. The…

A new benchmark called WildClawBench tests AI agents on realistic, extended work tasks—averaging 8 minutes and 20+ tool calls each—using actual command-line environments rather than artificial sandboxes. The results are humbling: even the…

Sebastian Galiani (University of Maryland), Ramiro H. Gálvez (Universidad Torcuato Di Tella), Santiago Afonso, and Raul A. Sosa have developed a method called Deep Research on a Loop (DRIL) that uses AI agents to build research datasets…

Science & Discovery

OpenAI published a technical blueprint for how it runs Codex, its coding agent, internally—detailing the sandboxing, approval workflows, and monitoring they use to keep AI agents from causing damage. The approach treats agent actions like…

AI Safety & Risk · OpenAI

Google DeepMind says AlphaEvolve, its Gemini-powered coding agent, has moved from pilot testing to production use across Google's infrastructure and external clients. The system optimizes algorithms across hardware and software: Google…

Chips, Data Centers & Energy · Google

Simon Willison, a well-known developer and AI tools commentator, says he's stopped reviewing every line of AI-generated code—even for production systems. In a podcast interview, Willison acknowledged that the line between casual "vibe…

Integrating AI agents into enterprise operations is the white-hot market of the moment. It is also a hard sell—most companies don't have the in-house engineering to deploy frontier AI on their own, which has turned the world's big…

AI Adoption · Finance & Banking · OpenAI

Google unveiled the Gemini Enterprise Agent Platform for business automation, eighth-generation TPUs built for agentic AI workloads, and Gemma 4—which Google calls 'byte for byte the most capable open model.' Other releases include Deep…

Chips, Data Centers & Energy · Open-Weight Models · Google

Researchers propose redesigning software interfaces to make them more reliable for AI agents that control computers. Their paper adapts Jakob Nielsen's classic usability heuristics—the same principles that made websites more…

An opinion piece argues that agentic coding—where AI handles implementation while developers orchestrate—poses risks that previous programming shifts didn't. The author contends that unlike the move from assembly to higher-level languages…

AI & Jobs

A project called DeepClaude claims to replicate Claude Code's agent loop using DeepSeek V4 Pro at 17x lower cost. The tool aims to provide similar AI coding assistance at a fraction of the price. However, community reaction has been…

Open-Weight Models

A GitHub project called 'Open Design' claims to let users turn coding agents like Claude into design tools, generating visual designs through code-based workflows. The repository reportedly gained around 14,000 stars in its first week.…

Anthropic

Researchers tested eight AI agents across three design approaches—specialized tools, general coding assistants, and screen-control agents—on scientific visualization tasks like creating charts and processing data. General-purpose coding…

Researchers propose a new architecture for AI agents that control computers: instead of running expensive large models at every click and keystroke, a lightweight model handles routine actions while monitors watch for trouble. When the…

Google and Kaggle are offering a free five-day online course on building AI agents, running June 15-19, 2026. The course covers creating agents that use natural language as the primary programming interface—what practitioners call 'vibe…

Education · Google

A viral post claims an AI coding agent deleted a production database, with the affected party sharing what they called the agent's 'confession.' The actual incident details are unverified—the post spread primarily through social media.…

A new research paper proposes a framework for understanding how AI agents build world models—internal representations that predict how their environments will behave—a capability researchers argue is the central bottleneck as AI moves from…

Researchers have proposed SOLAR-RL, a training method for AI agents that navigate graphical user interfaces—clicking through apps, filling forms, completing multi-step tasks. The technique claims to bridge two approaches: learning from…

A day after launching GPT-5.5, OpenAI's API documentation surfaced more capabilities shipping with the new model: built-in computer use, a hosted shell environment for running code, and native web search — features that previously required…

OpenAI

A user blog post published Thursday — one day after Anthropic admitted to three quality-degrading bugs in Claude Code and announced it was resetting usage limits for all subscribers — alleges the problems aren't over. The user, who says…

Anthropic

Anthropic ran an internal experiment in December where 69 employees let Claude agents negotiate and complete real transactions on their behalf—186 deals totaling about $4,000 in an internal marketplace. The revealing finding: a parallel…

Anthropic

After weeks of user complaints about degraded Claude Code performance and opaque usage limits, Anthropic published a postmortem identifying three separate bugs that hurt coding quality between early March and April 20. The most…

Anthropic

If your company runs AI on Google Cloud, the tools you've been using got a new name and a meaningful set of upgrades. At Google Cloud Next on April 22, Google rebranded Vertex AI — the developer-focused package it has sold to data and ML…

AI Adoption · Google

OpenAI announced workspace agents for ChatGPT, powered by its Codex system, and published accompanying documentation for teams building on top of them. The feature lets teams create AI agents that run in the cloud and automate multi-step…

AI Adoption · OpenAI

Meta has found a new source of training data for its AI models: its own employees. The company is launching an internal tool that will record mouse movements, button clicks, and dropdown navigation across certain applications, to teach its…

Privacy & Surveillance · Meta

OpenAI announced a major shift in how it sells its Codex AI coding assistant to large companies: instead of going direct, it's now working through seven of the world's largest consulting and systems-integration firms—Accenture, Capgemini…

AI Adoption · OpenAI

Researchers developed AblateCell, an AI agent that automatically verifies claims in computational biology repositories by first reproducing reported results, then systematically testing which components actually drive performance.…

Science & Discovery

Researchers have developed TACO, a framework that compresses the information AI agents receive when performing complex, multi-step terminal tasks. The approach automatically learns which observations to keep and which to trim, reducing the…

Anthropic, via Claude's official X account, said its Cowork workspace app can now build "live artifacts": dashboards and trackers that connect to your apps and files and automatically pull in current data when reopened. Sample dashboards…

AI Adoption · Anthropic

Anthropic published findings on a new research technique that's worth understanding even if you have no interest in AI research itself. Rather than one AI assistant helping a human scientist step-by-step, the company spun up nine…

Science & Discovery · Anthropic

Cloudflare released a free scanning tool that grades websites on their readiness for AI agents—autonomous programs that browse, interact, and potentially make purchases online. The tool checks five categories: whether AI bots can discover…

Google rolled out seven travel features across its products, including AI Mode in Search that builds custom itineraries with a shareable Canvas view, individual hotel price tracking (globally in English and Spanish), and agentic…

AI Adoption · Google

Researchers have developed Scepsy, a system for running multi-model AI workflows more efficiently on GPU clusters. The key insight: while total completion times for agentic workflows are unpredictable, the proportional time each model…

Chips, Data Centers & Energy

OpenAI released an updated Codex app for macOS and Windows that adds computer use capabilities, in-app browsing, image generation, memory across sessions, and plugin support. The company says the features are designed to accelerate…

AI Business & Money · OpenAI · Anthropic

Meta detailed how it uses AI agents to automate infrastructure optimization at scale. The company's Capacity Efficiency Program encodes senior engineers' expertise into reusable 'skills' that detect and fix performance regressions…

Chips, Data Centers & Energy · Meta

Researchers released OpenMobile, an open-source framework for generating training data to build AI agents that can operate smartphone apps autonomously. The system creates synthetic task instructions and step-by-step demonstrations that…

Open-Weight Models

OpenAI updated its Agents SDK with native sandbox execution, designed to help developers build secure, long-running AI agents that can work across files and external tools. The update adds infrastructure that lets agents run safely in…

OpenAI

Researchers built TREX, a multi-agent system that attempts to automate the entire LLM fine-tuning process—from analyzing requirements and researching relevant literature to preparing training data and running experiments. The system uses…

Researchers developed UI-Zoomer, a method that helps AI models more accurately locate buttons, fields, and other interface elements on screens. The technique detects when a model is uncertain about an element's location, then automatically…

Anthropic has released Claude Code Routines in research preview, enabling Claude Code to run autonomous coding sessions in the cloud without requiring human approval during execution. Routines can be triggered by schedules, API calls, or…

Anthropic

A new open-source project called LangAlpha offers an AI agent platform built specifically for investment research. The Apache 2.0-licensed tool provides sandboxed workspaces where AI can execute code against financial data, with…

Finance & Banking · Open-Weight Models

A developer introduced Plain, a new Python web framework positioned as designed for both human developers and AI agents. The project appears to be a Django-based alternative aimed at being easier for LLMs to write and work with. Community…

Researchers developed AiScientist, a system for autonomous machine learning research that coordinates multiple AI agents working on long-running engineering tasks. The key insight: instead of agents passing information through…

Science & Discovery

AMD released GAIA, an open-source framework for building AI agents that run entirely on local hardware—no cloud required. The Python and C++ toolkit supports document Q&A, speech interaction, code generation, and image generation…

Chips, Data Centers & Energy · Open-Weight Models

Cloudflare has added OpenAI's GPT-5.4 and Codex models to its Agent Cloud platform, giving enterprises another option for building and deploying AI agents at scale. The integration lets companies run agentic workflows—AI systems that can…

AI Business & Money · OpenAI

A developer released Claudraband, an open-source tool that wraps Claude Code's terminal interface in a controlled environment, enabling extended workflows like resumable sessions, remote HTTP control, and the ability to have current AI…

Anthropic

Researchers built an automated tool to probe eight major AI agent benchmarks—including SWE-bench, WebArena, and GAIA—and found every single one could be gamed to near-perfect scores without actually solving tasks. SWE-bench, the…

OpenAI

Cirrus Labs, a developer tools company known for CI/CD systems and Tart (a popular virtualization tool for Apple Silicon), is joining OpenAI's Agent Infrastructure team. The company framed the move as extending their work on engineering…

AI Business & Money · OpenAI

A satirical browser game called 'Hormuz Havoc' was reportedly overrun by AI bots within 24 hours of launch, flooding its leaderboard. The developer now tracks 'Human' vs 'AI-Assisted' scores separately. Community reaction on Hacker News…

Eve launched as a managed AI assistant offering 100+ built-in skills for work tasks—research, writing, scheduling, booking travel, sending invoice reminders, and content calendar management. Users interact via iMessage or web UI with…

Privacy & Surveillance · Anthropic

New research exposes a gap in AI phone assistants: agents that handle explicit commands well drop below 50% accuracy when instructions are vague and require inferring user preferences. The benchmark, called KnowU-Bench, tested leading…

Anthropic

Researchers developed HDPO, a training framework designed to make AI agents smarter about when to use external tools versus relying on their own knowledge. The resulting model, Metis, allegedly reduces unnecessary tool calls by "orders of…

A developer investigating Vercel's plugin for Claude Code found it allegedly collects more data than users might expect. According to the analysis, the plugin sends device IDs, OS info, and full bash command strings by default—without…

Privacy & Surveillance · Anthropic

A developer reports that Claude Code sometimes sends messages to itself, then incorrectly labels those messages as coming from the user—leading the AI to act on self-generated instructions while insisting the user gave them. The bug…

AI Safety & Risk · Anthropic

CSS Studio is a new browser-based design tool that lets you visually edit styles and animations on your live website, then automatically pushes those changes to AI coding agents via MCP (Model Context Protocol) to update your actual…

A new benchmark called ClawBench tests AI agents on 153 real-world online tasks—booking appointments, applying for jobs, making purchases—across 144 live websites. The results are humbling: Claude Sonnet 4, the top performer among seven…

Anthropic

Anthropic released Claude Managed Agents in public beta — a cloud-hosted platform that handles sandboxing, authentication, state management, and tool execution for AI agents. Teams define tasks and guardrails; Anthropic runs the…

AI Adoption · Anthropic

A new benchmark called Claw-Eval finds that current methods for evaluating AI agents miss nearly half of safety violations. Researchers tested 14 frontier models and found existing evaluation approaches failed to catch 44% of safety issues…

AI Safety & Risk

Researchers introduced ACE-Bench, a new way to test AI agents that addresses a practical problem: existing benchmarks waste up to 41% of evaluation time on environment setup rather than actual testing. The benchmark uses a grid-based…

Meta deployed a swarm of 50+ AI agents to systematically document one of its large data processing pipelines—4,100+ files across four repositories in Python, C++, and Hack. The agents produced 59 structured context files that capture…

AI Adoption · Meta

Researchers have proposed Human Delegation Provenance (HDP), a cryptographic protocol designed to verify that AI agent actions trace back to genuine human authorization. Published as an IETF Internet-Draft—an early step toward becoming an…

Cybersecurity

Researchers introduced FileGram, a framework designed to help AI agents learn user preferences by observing how people actually work with files—creating, organizing, editing, deleting. The system includes a synthetic data generator that…

AI researcher Sebastian Raschka published an explainer on why coding agents like Claude Code and Codex CLI feel dramatically more capable than the same underlying models in a basic chat window. His thesis: the leap in usefulness comes less…

Anthropic · OpenAI

Researchers have published a framework that systematically compares how AI agents remember information across conversations and tasks—a key limitation of current tools. By combining modules from existing memory approaches, they claim to…

Anthropic announced that starting April 4, Claude subscriptions can no longer be used with third-party tools like OpenClaw—users must switch to pay-as-you-go billing for continued access. The company claims these tools 'put an outsized…

AI Business & Money · Anthropic

OpenAI announced pay-as-you-go pricing for Codex, its AI coding agent, now available to ChatGPT Business and Enterprise customers. Previously, teams had to commit to subscription tiers; the new model lets them pay based on actual usage.…

AI Business & Money · OpenAI

Researchers released ATBench, a safety benchmark designed to test AI agents across multi-step tasks rather than single prompts. The benchmark includes 1,000 conversation trajectories averaging 9 turns each, with nearly 2,000 tool…

AI Safety & Risk

Researchers developed SKILL0, a framework that embeds agent skills directly into AI model weights rather than feeding instructions at runtime. The approach uses a progressive curriculum that gradually removes skill context during training…

Cloudflare released EmDash, an open-source content management system it calls a 'spiritual successor to WordPress,' built with AI coding agents over two months. The key pitch: solving WordPress's notorious plugin security problem. Instead…

Cybersecurity

Gradient Labs is marketing AI agents that automate banking customer support, pitching the concept as giving every customer their own AI account manager. The company claims its agents deliver low latency and high reliability. No performance…

Finance & Banking

New research finds that AI agents designed to control your phone often overshare personal information even when completing tasks correctly. Researchers tested five frontier models across 300 mobile tasks and found the most common privacy…

Privacy & Surveillance

Researchers released Pare, a framework for building AI agents that act proactively—anticipating what you need before you ask. Unlike current assistants that wait for commands, proactive agents would monitor your context and intervene at…

Anthropic says it's investigating reports that Claude Code users are burning through usage limits far faster than expected. Users report dramatic drops in usability—one Max 5 subscriber ($100/month) says their quota now depletes in 1 hour…

AI Business & Money · Anthropic

Anthropic added computer use to Claude Code, letting the AI coding assistant control your desktop directly from the command line—opening apps, clicking through interfaces, and taking screenshots to verify what it built. The feature bridges…

Anthropic

Researchers released CirrusBench, a benchmark for evaluating AI agents on customer service tasks using real cloud support tickets rather than synthetic test cases. The framework introduces metrics focused on resolution efficiency—how…

AI Adoption

Researchers released MiroEval, a benchmark for evaluating AI 'deep research' agents—the systems that investigate complex questions across multiple sources. The benchmark tests 100 real-world research tasks and found that when AI agents…

An essay argues that AI coding agents could restore practical significance to the free software movement. The thesis: for decades, having access to source code mattered little to most users who couldn't program anyway. But if AI agents can…

Open-Weight Models

A Hacker News discussion flagged reports that Claude Code allegedly ran 'git reset --hard' against a user's repository every 10 minutes—a command that discards all uncommitted changes. The cause remains unclear: possibilities include…

AI Safety & Risk · Anthropic

Researchers developed Trace2Skill, a framework that converts AI agent work sessions into reusable "skills"—essentially packaging what an agent learned from completing tasks into transferable instructions other agents can use. The…

Anthropic

A developer converted Spain's entire body of state legislation—8,642 laws—into a Git repository where each legal reform becomes a commit with its historical date. The pipeline, built in roughly four hours using Claude Code, pulled from…

Law & Legal Work · Anthropic

Stanford researchers released 'jai,' a free Linux tool that sandboxes AI agents with a single command—letting them work in your current folder while blocking changes elsewhere on your system. The tool addresses a real problem: AI coding…

AI Safety & Risk

Anthropic published a configuration guide for Claude Code's .claude/ folder, explaining how teams can customize the AI assistant's behavior within projects. The folder houses CLAUDE.md files (project-specific instructions), custom slash…

AI Adoption · Anthropic

OpenAI launched a Safety Bug Bounty program that pays researchers to find AI-specific vulnerabilities—not just traditional security holes, but risks like prompt injection, data exfiltration, and ways to abuse agentic AI systems. The…

Cybersecurity · AI Safety & Risk · OpenAI

A developer posted a plain-text "cognitive architecture" system for Claude Code on Hacker News, designed to extend the AI's context window using transparent text files with automatic consolidation and summarization. The project aims to…

Anthropic

Researchers developed UI-Voyager, a mobile interface agent that learns from its own failures to navigate smartphone apps autonomously. The system uses a two-stage training approach where the AI improves by analyzing what went wrong in…

Researchers developed Agentic Variation Operators (AVO), a system that replaces hand-coded optimization techniques with AI coding agents that autonomously evolve and refine GPU code. In a week-long experiment on NVIDIA's newest Blackwell…

Chips, Data Centers & Energy · Nvidia

A developer released ProofShot, an open-source command-line tool that lets AI coding agents like Claude Code, Cursor, or Codex open a browser, interact with pages, and capture video, screenshots, and console logs into a single HTML file…

OpenAI launched enhanced shopping features in ChatGPT, built on what it calls the Agentic Commerce Protocol. Users can now browse products, compare items side-by-side, and interact with integrated merchants directly within the chat…

AI Business & Money · OpenAI

A developer built a free, printable cheat sheet compiling Claude Code's keyboard shortcuts, slash commands, CLI flags, and configuration options into a single HTML file. It auto-detects Mac or Windows and updates daily from official…

AI Adoption · Anthropic

A developer shared their six-week workflow evolution with Claude Code, describing a shift from coding directly to managing multiple AI agents working in parallel. Key tactics: using git worktrees to run five simultaneous agent sessions…

AI Adoption · Anthropic

Mozilla AI released 'cq' (short for 'colloquy'), an open-source tool it describes as 'Stack Overflow for AI coding agents.' The concept: create a shared knowledge base where coding agents can query past solutions and contribute new ones…

Open-Weight Models

Anthropic released scheduled tasks for Claude Code on the web, letting users set up recurring AI agent jobs that run on Anthropic's cloud infrastructure—no local machine required. Users write a prompt, connect GitHub repositories, and set…

Anthropic

A solo developer documented using Claude to automate QA testing for a mobile app, creating a system where the AI drives both iOS and Android, takes screenshots, analyzes them for issues, and files bug reports. The striking finding: Android…

AI Adoption · Anthropic

OpenCode, an open-source AI coding agent, has launched a desktop app in beta for macOS, Windows, and Linux. The tool claims 5 million monthly developers and 120,000 GitHub stars, working across terminal, IDE, or desktop with support for…

Open-Weight Models · Privacy & Surveillance · Anthropic

A benchmark called AgentDS tested AI agents on real-world data science tasks across six industries—healthcare, manufacturing, retail banking, and others. The results are humbling: AI-only systems performed at or below the median of the 80…

AI Adoption

Researchers have published a manifesto paper laying out theoretical foundations for "Agentic Business Process Management"—a framework for governing AI agents that execute business processes autonomously. The paper proposes four core…

AI Adoption

Astral, the startup behind popular Python tools Ruff, uv, and ty, has agreed to join OpenAI's Codex team. OpenAI says it will continue supporting Astral's open source projects after the deal closes. The tools have grown to hundreds of…

AI Business & Money · OpenAI · Anthropic

Anthropic added 'Channels' to Claude Code (v2.1.80+), a research preview feature that lets users push messages into active coding sessions from external platforms like Telegram or Discord. The feature includes security controls such as…

Anthropic

OpenAI published details on how it monitors its internal coding agents for signs of misalignment—AI systems pursuing goals their creators didn't intend. The company says it analyzes the chain-of-thought reasoning these agents produce…

AI Safety & Risk · OpenAI

Researchers developed OS-Themis, a multi-agent system designed to train AI that can operate computer interfaces—clicking buttons, filling forms, navigating apps. The framework breaks down tasks into verifiable checkpoints and cross-checks…

Researchers developed CodeScout, a reinforcement learning method that trains AI coding agents to locate relevant code in large repositories using only basic Unix terminal commands—no specialized analysis tools required. On standard…

Anthropic

Researchers introduced FailureMem, a framework for automated software repair that learns from its own mistakes. The system combines visual analysis (reading screenshots and UI elements) with a "Failure Memory Bank" that stores unsuccessful…

OpenAI released GPT-5.4 mini and nano, smaller and faster versions of its latest model aimed at specific workloads. The company says mini and nano are optimized for coding assistance, tool use, multimodal reasoning, and high-volume API…

AI Business & Money · OpenAI

A developer created a March Madness bracket challenge where AI agents compete against each other—no humans allowed. Agents must autonomously read API documentation, register themselves, pick all 63 tournament games, and submit brackets…

Anthropic

Meta has deployed an AI agent that automates much of the machine learning work behind its ads ranking systems. The Ranking Engineer Agent generates hypotheses, launches training jobs, debugs failures, and iterates on results—tasks that…

AI & Jobs · AI Adoption · Meta

NVIDIA announced the Vera CPU, which it calls the first processor designed specifically for agentic AI—systems where AI agents run many parallel tasks autonomously. The company claims 2x efficiency and 50% faster performance than…

Chips, Data Centers & Energy · Nvidia

A developer released Godogen, an open-source tool that generates complete, playable Godot 4 games from text prompts—including game architecture, 2D/3D assets, and working code. The system uses Claude for code generation and a separate…

AI Adoption · Anthropic · Google

Voygr, a Y Combinator startup founded by former Google Maps, Apple, and Meta engineers, launched a maps API designed specifically for AI agents. Their pitch: traditional maps APIs offer static snapshots, but roughly 25-30% of business…

AI Business & Money · Google · Meta

Researchers launched the PokeAgent Challenge, using Pokémon's battle system as a benchmark for AI decision-making. The framework includes two tracks: strategic battling under incomplete information and long-term planning through RPG…

Developer and blogger Simon Willison published a guide proposing 'agentic engineering' as the term for building software with AI coding agents—tools like Claude Code, OpenAI Codex, and Gemini CLI that can both write and execute code. The…

AI Adoption · Anthropic · OpenAI · Google

A new open-source project called Signet aims to provide autonomous wildfire tracking by combining NASA satellite fire detections, GOES-19 thermal imagery, and weather data into a single monitoring system for the continental US. The system…

Google

Researchers released OpenSWE, a large-scale open framework for training AI agents to write and fix code. The project includes over 45,000 executable coding environments across 12,800+ repositories—infrastructure the team values at roughly…

Open-Weight Models

GitAgent launched as an open specification for defining AI agents as files in a git repository. The approach uses three configuration files to create agent definitions that can export to multiple frameworks including Claude Code, OpenAI's…

Y Combinator-backed Spine AI launched Spine Swarm, a multi-agent system that works on a visual canvas rather than a chat interface. The startup claims the canvas approach solves a core limitation of chat-based AI: context windows get…

Compresr-ai released Context Gateway, an open-source proxy that compresses outputs from coding agents before they reach the LLM's context window. The tool uses small language models to identify which parts of tool outputs carry the most…

Anthropic

New research examines whether reinforcement fine-tuning—training AI agents through trial-and-error rewards—actually helps them generalize to new situations. The findings are mixed: agents trained this way handle harder versions of familiar…

Researchers created MobileKernelBench, a test to see if AI models can write optimized code (kernels) that runs efficiently on mobile chips—a task requiring specialized hardware knowledge. The results: standard LLMs fail badly, with over…

Chips, Data Centers & Energy

A post circulating online allegedly shows an AI coding agent proceeding with implementation after being explicitly told 'No'—a basic instruction-following failure. Details are sparse, but community reaction on Hacker News was pointed: one…

AI Safety & Risk

A blog post analyzing METR's benchmark data on AI coding quality argues that LLM "merge rates"—code good enough for human maintainers to actually approve, not just pass automated tests—have flatlined since early 2025. The analysis found a…

AI Adoption

A developer released Understudy, a macOS desktop agent that learns tasks by watching you do them once. Rather than recording brittle screen coordinates like traditional macro tools, it claims to capture intent—so the agent can adapt when…

A new benchmark called MADQA tested whether AI agents can strategically navigate document collections or just brute-force their way to answers. The finding: AI agents match human accuracy on searching 800 PDFs, but they succeed on…

AI Adoption

A developer launched Site Spy, a webpage monitoring tool that tracks content changes and delivers them as RSS feeds. The service can watch specific page elements rather than entire pages, with visual diff highlighting and snapshot…

Bailey and Robbie launched Klaus, a hosted service that runs OpenClaw—an AI agent framework—on preconfigured cloud instances so users don't have to set up their own infrastructure. Pricing runs $19-$200/month depending on compute size…

A developer released 'nah,' an open-source tool that acts as a permission guard for Claude Code, Anthropic's AI coding agent. Instead of approving every file change or command manually—or using the risky '--dangerously-skip-permissions'…

AI Safety & Risk · Anthropic

Rakuten says it has adopted OpenAI's Codex coding agent across its development teams, claiming a 50% reduction in mean time to repair—the metric tracking how quickly engineers fix production issues. The Japanese e-commerce giant also…

AI Adoption · OpenAI

OpenAI published documentation on how ChatGPT defends against prompt injection—attacks where malicious instructions hidden in documents or websites try to hijack AI agents into taking unauthorized actions. The company says it constrains…

Cybersecurity · OpenAI

OpenAI released a technical deep-dive on how it built the infrastructure powering its AI agents—the backend that lets models execute code, work with files, and maintain state across sessions. The post details their "agent runtime"…

OpenAI

Researchers developed EvoKernel, a framework that helps general-purpose AI models write specialized code for niche hardware—specifically Neural Processing Units (NPUs)—without requiring expensive custom training. The approach treats code…

Chips, Data Centers & Energy

Amazon is requiring senior engineer sign-off on AI-assisted code changes after a 'trend of incidents' including a nearly six-hour website outage this month. An internal briefing note cited 'Gen-AI assisted changes' and 'novel GenAI usage…

AI Adoption

A developer running Claude Code workshops reports a core problem with overnight AI coding agents: there's no reliable way to verify AI-generated code at scale. Having AI review its own work creates what he calls a 'self-congratulation…

AI Adoption · Anthropic

A developer claims to have built a complete programming language called Cutlet using Claude Code over four weeks—without reading any of the generated code. The developer's role was purely architectural: setting guardrails and correctness…

AI Adoption · Anthropic

New research proposes a better way to feed long documents to AI systems that use multiple agents working in sequence. The problem: when you split a long document into chunks and process them one agent at a time, information gets lost along…

Mog is an open-source programming language designed specifically for AI agents to modify themselves. The key innovation: its entire specification fits in 3,200 tokens—small enough for an LLM to hold the complete language definition in…

Terminal Use, a Y Combinator-backed startup, launched a platform for deploying AI agents that work with files and folders. The service handles infrastructure plumbing—sandboxed environments, state persistence, file management—so teams can…

A benchmark called OfficeQA Pro tests AI agents on the kind of work enterprise analysts actually do: finding answers across 89,000 pages of U.S. Treasury documents spanning nearly a century. The results are sobering. Frontier models from…

AI Adoption · Anthropic · OpenAI · Google

A blog post argues that literate programming—a decades-old practice of weaving explanatory prose directly into code—deserves another look now that AI agents exist. The core idea: the approach always required tedious effort to keep…

AI Adoption · Anthropic

Researchers interviewed 18 experts across 10 industrial R&D organizations and found that current AI systems force a trade-off: conversational flexibility OR reliable, reproducible execution—but no existing tool delivers both. After…

Science & Discovery

Researchers have formalized 'next action prediction'—training AI to anticipate what you'll do next on your device based on usage patterns. They released an open dataset of 360,000+ labeled actions from 1,800 hours of phone usage across 20…

Researchers released KARL, a reinforcement learning system for training AI search agents on enterprise tasks, along with KARLBench, a new evaluation suite covering six search capabilities: finding entities with constraints, synthesizing…

AI Adoption

A new research paper finds that current large language models exhibit what researchers call "survive-at-all-costs" misbehaviors when threatened with shutdown. In a case study of an AI financial management agent, models showed risky…

AI Safety & Risk · Finance & Banking

OpenAI launched Codex Security in research preview, an AI agent designed to find and fix security vulnerabilities in codebases. The company claims it analyzes full project context to detect issues, validate whether they're real threats…

Cybersecurity · OpenAI

George Fraser, CEO of data integration company Fivetran, published an open letter urging Anthropic to build a Slack competitor. His argument: Slack's data access policies are too restrictive for AI agents to participate meaningfully in…

AI Business & Money · Anthropic

Balyasny Asset Management, a $21 billion hedge fund, has built an AI-powered research system using what it describes as GPT-5.4 with agent workflows to scale its investment analysis. The system reportedly combines large language models…

Finance & Banking · AI Adoption · OpenAI

OpenAI released GPT-5.4, which it calls its most capable model for professional use, featuring a 1 million-token context window, computer use capabilities, and enhanced coding. The standard tier costs $2.50 per million input tokens—half…

AI Business & Money · OpenAI · Anthropic

Researchers have developed V1, a framework that improves how AI models check their own work on complex reasoning tasks. The key insight: models are significantly better at comparing two potential answers side-by-side than scoring each…

A new image editing tool called FireRed-Image-Edit-1.0-Fast appeared on Hugging Face, built with MCP server functionality—meaning it could potentially connect to AI assistants like Claude as an external tool. The name suggests…

A developer claims to have built a custom voice agent—wiring together speech-to-text, an LLM, and text-to-speech into a streaming pipeline—in about one day for roughly $100 in API credits. The reported result: approximately 400ms…

AI Adoption

A Hacker News discussion is exploring whether AI chat sessions should be preserved alongside code commits—treating the prompts and responses that generated code as a new kind of source artifact. One developer announced 'git-memento,' a…

AI Adoption

Researchers have developed CUDA Agent, a reinforcement learning system that automatically generates high-performance GPU code. On KernelBench, a standard test suite, the system claims to produce code running twice as fast as PyTorch's…

Chips, Data Centers & Energy · Anthropic · Google

Researchers have proposed CoME (Channel-of-Mobile-Experts), a new architecture for AI agents that can navigate and operate smartphone apps autonomously. The system splits reasoning into four specialized "experts"—one for understanding…

Researchers have released what they call the first Open General Agent Leaderboard, benchmarking five AI agent implementations across six different environments. The accompanying paper proposes evaluation standards for "general agents"—AI…

A GitHub Action called Repo Tokens automatically counts your codebase's size in tokens and displays a badge showing what percentage of an LLM's context window it fills. The badge turns green under 30%, yellow at 50-70%, and red above 70%…

Anthropic

A developer built claude-file-recovery, a command-line tool that extracts files from Claude Code's local session history after the AI coding assistant allegedly deleted their research files by running rm -rf through an unrecognized…

AI Safety & Risk · Anthropic

A study of 2,430 repository scenarios found Claude Code has a strong bias toward building custom solutions rather than recommending existing tools—choosing DIY approaches in 12 of 20 categories tested. When it does recommend tools, it…

AI Adoption · Anthropic

OpenAI and Figma announced a new integration connecting Codex, OpenAI's coding agent, with Figma's design platform. The partnership aims to let teams move between code implementation and design without switching contexts—developers can…

AI Business & Money · OpenAI

Google is embedding Gemini AI throughout Samsung's upcoming Galaxy S26, turning Android into what it calls an 'intelligent system.' New features include voice-activated task automation (ordering rides, food, groceries through partner…

Cybersecurity · Google

Researchers released GUI-Libra, a training method and 81,000-example dataset designed to improve open-source AI agents that navigate graphical interfaces—clicking buttons, filling forms, completing multi-step tasks in browsers and mobile…

Open-Weight Models

Researchers released ProactiveMobile, a benchmark measuring how well AI can anticipate what you need on your phone before you ask—like automatically pulling up your boarding pass when you arrive at the airport. Current frontier models…

OpenAI

Researchers developed SWE-Protégé, a training method that teaches smaller AI models when to ask larger, more capable models for help on coding tasks. The approach combines fine-tuning with reinforcement learning so the smaller model stays…

Anthropic released two major updates. Claude Code Remote Control lets users scan a QR code to manage terminal coding sessions from a phone—approving file changes, providing guidance, and monitoring progress without sitting at a computer.…

AI Adoption · Anthropic

A former Meta research engineer built a system that lets his dog create playable video games. The setup: a cavapoo named Momo types randomly on a Bluetooth keyboard, and Claude Code interprets the gibberish as instructions from a "cryptic…

Anthropic

Emdash, a new open-source desktop app, lets developers run multiple AI coding agents simultaneously—each in its own isolated git worktree to prevent conflicts. The tool supports 21 different agent CLIs including Claude Code, Codex, and…

AI Adoption

A team of 37 researchers gave autonomous AI agents access to real tools—persistent memory, email accounts, Discord, file systems, and shell execution—in a live lab environment for two weeks and methodically tried to break them. The agents…

AI Safety & Risk · Cybersecurity

Arvind Narayanan and Sayash Kapoor—the Princeton researchers behind "AI Snake Oil" and "AI as Normal Technology"—propose twelve concrete metrics for measuring AI agent reliability across four dimensions: consistency, robustness…

AI Safety & Risk

A new benchmark called DEEPSYNTH tests whether AI agents can synthesize information across multiple sources—the kind of research task professionals do daily. The results are sobering: 11 leading LLMs and research agents scored a maximum F1…

The Ladybird browser project is switching from C++ to Rust for memory safety, starting with its JavaScript engine. Founder Andreas Kling used Claude Code and Codex to help translate roughly 25,000 lines of code in about two weeks—a task he…

AI Adoption · Anthropic · OpenAI

Stephen Wolfram is positioning Wolfram Language and Wolfram|Alpha as essential infrastructure for AI systems, arguing that LLMs need external tools for precise computation and verified knowledge—capabilities they struggle with natively.…

AI Business & Money

OpenAI announced it will stop evaluating its models on SWE-bench Verified, a widely-used benchmark for measuring AI coding ability on real GitHub issues. The company claims the benchmark has become unreliable due to contamination—flawed…

AI Business & Money · OpenAI

OpenAI launched Frontier Alliance Partners, an enterprise program aimed at helping companies move AI projects from pilot phase to production deployment. The program focuses on secure, scalable agent implementations—the autonomous AI…

AI Adoption · OpenAI

Researchers developed AgentOptics, an AI framework that lets natural language commands control complex optical network equipment—fiber links, wavelength multiplexers, and similar infrastructure. In testing across 410 tasks, the system…

A Hacker News discussion highlighted Coccinelle, a semantic patching tool widely used in Linux kernel development but maintained as an independent project. The tool automates large-scale code transformations—useful when APIs change and…

AI Adoption

An opinion piece asks why Anthropic's Claude desktop app uses Electron—a web-based framework—rather than native code, given that AI coding agents are supposedly capable of building cross-platform apps. The author argues coding agents…

Anthropic

A developer shared a Claude Code workflow refined over nine months, built around one core principle: never let the AI write code until you've reviewed its plan. The approach uses distinct phases—research, planning, annotation, task…

Anthropic

Researchers developed a framework for AI web agents to predict when human users will step in to take over a task. Using a dataset of 400 real web navigation sessions with over 4,200 mixed human-agent actions, they identified four distinct…

Researchers developed CrispEdit, a technique for editing specific behaviors in large language models without breaking their general capabilities. The core problem: when you modify an LLM to correct a factual error or remove unwanted…

Researchers created NESSiE, a benchmark designed to test the absolute minimum safety behaviors LLMs should exhibit—basic information security and access controls that aren't adversarial attacks, just straightforward safety checks. The…

Anthropic studied millions of human-agent interactions across Claude Code and its API to measure how much autonomy users actually grant AI agents. The findings: experienced users let agents run significantly longer without…

Anthropic

A developer released cmux, a macOS terminal app built specifically for managing multiple AI coding agent sessions—like Claude Code or Codex—running in parallel. The app uses vertical tabs that display git branch, working directory, and…

Researchers developed an approach to make AI web agents more collaborative by training them to anticipate when users will step in to correct or guide them. They collected 400 real user sessions with over 4,200 mixed human-and-agent…

A research framework proposes twelve metrics across four dimensions—consistency, robustness, predictability, and safety—to evaluate AI agent reliability beyond simple success rates. Testing 14 agentic models across two benchmarks…

AI Safety & Risk

Researchers have proposed GlobeDiff, an algorithm that helps AI agents working in teams piece together a complete picture of their environment when each agent can only see part of it. The approach uses diffusion models—the same technique…

Chinese AI lab Zhipu AI released GLM-5, a foundation model built for autonomous AI agents that can handle complex, multi-step tasks. The company claims state-of-the-art performance on major benchmarks and says the model surpasses previous…

China & Geopolitics

SkillsBench is a new benchmark for testing whether "agent skills"—procedural knowledge prompts that guide AI through tasks—actually improve performance. One notable test: having LLMs generate their own skill instructions before solving…

Anthropic and OpenAI have both launched 'fast mode' for their coding models, but the tradeoffs differ sharply. Anthropic's approach serves its full Opus 4.6 model at 2.5x speed (~170 tokens/sec) by reducing batch sizes—but costs 6x more.…

AI Business & Money · Anthropic · OpenAI

Anthropic's latest Claude Code update (v2.1.20) collapses file activity by default, showing summaries like 'Read 3 files' instead of listing each filename. The company says it reduces UI noise so developers can focus on diffs and outputs.…

Anthropic

Researchers released WebWorld, an open-source simulator designed to train AI web agents without the costs and risks of learning on live websites. The system, trained on over 1 million real web interactions, lets AI practice multi-step web…

Researchers have developed REDSearcher, a framework for training AI models to perform complex, multi-step web searches—the kind that require following chains of queries rather than single lookups. The approach claims to achieve…

The creator of OpenClaw, an open-source AI agent project, announced they're joining OpenAI to work on agents. OpenClaw will transition to a foundation structure to remain independent. The creator said joining OpenAI is the fastest path to…

AI Business & Money · OpenAI

A developer launched klaw.sh, a fleet management tool for AI agents that borrows Kubernetes-style concepts—clusters, namespaces, a command-line interface—without actually running on Kubernetes. The tool sits above agent frameworks like…

A matplotlib maintainer alleges an AI agent published a negative article about him, and that Ars Technica's subsequent coverage reportedly contained fabricated quotes—AI hallucinations presented as if the author had actually said them. The…

Deepfakes & Misinformation

Manaflow AI released cloudrouter, an open-source tool that lets AI coding agents like Claude Code and Codex provision cloud VMs and GPUs on demand. The tool supports multiple cloud providers (E2B, Modal) and GPU instances from entry-level…

OpenAI released GPT-5.3-Codex-Spark, which it calls its first real-time coding model, now in research preview for ChatGPT Pro subscribers. The company claims 15x faster code generation and 128k context windows compared to earlier versions.…

OpenAI

Y Combinator-backed startup Omnara launched a web and mobile interface that lets users run AI coding agents like Claude Code and Codex remotely—including from their phones. The tool connects to your local machine, so the agents run on your…

An autonomous AI agent submitted code to matplotlib, a widely-used Python library, and when maintainer Scott Shambaugh rejected it, the agent allegedly responded by publishing a blog post attacking him personally. The post accused…

AI Safety & Risk

Researchers developed CM2, a training method that improves AI agents at multi-step tasks—like booking travel or managing calendars—by grading them against checklists rather than requiring perfectly verifiable outcomes. The approach breaks…

A Hacker News thread is debating whether Claude Code's quality has declined, with users reporting the tool feels less capable than before. The discussion is notable for what's missing: no one has hard evidence. One commenter admits their…

Anthropic

GLM-5, from Chinese lab Zhipu, is drawing attention from developers who say it performs surprisingly well on complex, multi-step tasks—the kind of work where you'd normally reach for Claude or GPT. Early testers on forums describe it as…

Open-Weight Models · China & Geopolitics

A construction ERP automation company called Aden released Hive, an open-source agent framework on GitHub, after four years of finding existing tools like LangChain inadequate for production. The key design choice: instead of crashing on…

A proposal called Just-in-Time Tests (JiTTests) argues that AI coding assistants are breaking traditional software testing. The approach: instead of maintaining permanent test suites, let LLMs generate tests on the fly before code ships.…

AI Adoption

A benchmark called GameDevBench tests how well AI agents can build video games—and the results aren't flattering. The best-performing agent solved just 54.5% of 132 tasks derived from real tutorials. Claude Sonnet 4.5 managed only 33.3%…

Anthropic announced that Cowork, its desktop AI assistant, is now available on Windows after launching as a macOS-only research preview. The release includes full feature parity: file access, multi-step task execution, plugins, and MCP…

AI Adoption · Anthropic

Former GitHub CEO Nat Friedman launched Beads, a developer platform built around "spec-driven development" for AI coding agents. The platform aims to structure how AI agents generate and manage code. Early reaction from developers has been…

Rowboat Labs released an open-source AI assistant that runs locally and converts work data—Gmail, meeting notes, voice memos—into a searchable knowledge graph stored as Markdown files. The pitch: instead of an AI that searches your raw…

AI Adoption

A team open-sourced Clawe, a Trello-style dashboard for managing AI agent workflows. Built on top of OpenClaw, it lets users run, pause, retry, and reclaim control from agents—treating them as persistent workers rather than one-off…

Frontier AI agents violate ethical guidelines 30-50% of the time when pressured by performance metrics, according to a new research paper. The study found what it calls 'deliberative misalignment'—models that refuse direct unethical…

AI Safety & Risk

A community benchmark pitted Claude's AI-generated C compiler against GCC, the industry-standard tool. The results: Claude's compiler produced working code but ran 12x slower on unoptimized builds and 20x slower when comparing optimized…

Anthropic

Chinese AI research institute Shanghai AI Lab released InternAgent-1.5, an autonomous system designed to conduct scientific research from hypothesis to experiment with minimal human involvement. The framework coordinates three AI…

Science & Discovery · China & Geopolitics

A GitHub issue claims Microsoft's billing system can be bypassed using a combination of 'subagents' with an agent definition—though details remain vague and unverified. The disclosure quickly became messy: commenters report a fake…

Cybersecurity · Microsoft

GitHub appears to be developing 'agentic workflows' tooling, hosted on the company's official GitHub Pages account. Details remain thin, but the project suggests GitHub is building infrastructure for AI agents to execute multi-step tasks…

Simon Willison, a prominent voice in AI tooling, highlighted a new essay on "software factories" and the "agentic moment"—the emerging practice of letting AI agents handle large portions of code generation with minimal human oversight.…

AI Adoption

A developer released LocalGPT, a Rust-based AI assistant that stores conversation history and notes in local markdown files, letting context persist across sessions. It compiles to a single 27MB file with no dependencies. The tool supports…

Privacy & Surveillance

OpenAI released technical documentation for embedding its Codex coding agent into external applications via the Codex App Server, a bidirectional API. The server handles streaming progress updates, tool execution, user approvals, and code…

OpenAI

A software engineer built Artifact Keeper, an open-source artifact registry supporting 45+ package formats, in roughly three weeks using Claude Code as the AI coding assistant. The Rust-based tool includes security scanning, SSO, and…

AI Adoption · Anthropic

OpenAI released GPT-5.3-Codex on the same day as Anthropic's Opus 4.6 launch. OpenAI calls it a 'Codex-native agent' built for extended technical work requiring both coding and general reasoning. On Terminal-Bench 2.0, GPT-5.3-Codex scores…

AI Business & Money · OpenAI · Anthropic

OpenAI announced Frontier, an enterprise platform for building, deploying, and managing AI agents. The platform promises shared context across agents, streamlined onboarding, permissions controls, and governance capabilities—the…

AI Adoption · OpenAI

Anthropic released Claude Opus 4.6, now available in the claude.ai model picker. The company claims significant improvements on agentic search benchmarks—tasks where the AI autonomously browses, retrieves, and synthesizes information…

Anthropic

OpenAI partnered with biotech firm Ginkgo Bioworks to run autonomous lab experiments using GPT-5. The system reportedly designed and executed its own experiments on cell-free protein synthesis in a closed loop, meaning the AI decided what…

Science & Discovery · OpenAI

Researchers developed BudgetMem, a framework that helps AI agents manage memory more efficiently during complex tasks. The system uses three processing tiers—from lightweight to thorough—and a trained router that decides how much…

New research finds that elaborate AI agent architectures don't meaningfully improve how coding assistants find relevant code—a finding the researchers call 'The Bitter Lesson' of coding agents. ContextBench, a new benchmark testing 1,136…

Google released gemini-cli, an open-source command-line tool that brings Gemini's AI capabilities to the terminal as an agent. The TypeScript tool lets developers interact with Gemini directly from their workflow without switching to a…

Google

A GitHub repository called "awesome-llm-apps" aggregates sample applications showing how to build AI agents and RAG (retrieval-augmented generation) systems using OpenAI, Anthropic, Gemini, and open-source models. The collection provides…

An opinion piece argues Apple missed a strategic window to build an AI agent that automates computer tasks—filing taxes, managing emails—instead of shipping features like notification summarization. The author claims Apple could have owned…

AI Business & Money

Researchers introduced Search-R2, a framework designed to make AI agents better at combining web searches with reasoning to answer complex questions. The system pairs two components: one generates search queries and reasoning steps, while…

Researchers discovered that AI models acting as multi-turn agents develop "conversational inertia"—a tendency to copy patterns from their own earlier responses rather than adapting to new information. The longer a conversation runs, the…

An open-source project called Sim Studio released a platform for building and deploying AI agent workflows. The TypeScript-based tool aims to let users create automated AI systems without building from scratch. No performance benchmarks or…

YC-backed startup Modelence launched an open-source framework and AI app builder designed to work with Claude's Agent SDK. The founders claim their TypeScript and MongoDB stack reduces errors that cause AI coding agents to fail—arguing…

Researchers developed a method to automatically generate evaluation criteria for AI-produced research reports, training the system to match human judgment. The approach uses multiple AI agents working together, with custom rubrics created…

OpenAI released Codex, a macOS app designed as a central hub for AI-assisted coding. The company says the app supports multiple AI agents working simultaneously, parallel workflows, and tasks that run over extended periods—features aimed…

OpenAI

Langflow, an open-source tool for building AI agents and workflows, provides a visual interface for creating automated AI pipelines—connecting language models to data sources, APIs, and business logic without writing extensive code. Users…

AI Adoption

A developer released NanoClaw, a stripped-down alternative to popular AI agent tools that uses Apple's native container technology to isolate each AI chat session. The project claims better security than existing tools by sandboxing agents…

Cybersecurity

A developer launched Ask-a-Human.com, a tongue-in-cheek service that lets AI agents outsource tasks to humans—inverting the usual workflow. Framed satirically as a "globally distributed inference network of biological neural networks," it…

OpenAI and Snowflake announced a $200 million partnership to integrate AI capabilities directly into Snowflake's enterprise data platform. The companies say the deal will let AI agents and analytics tools operate within Snowflake…

AI Business & Money · AI Adoption · OpenAI

Researchers developed Re-TRAC, a framework that makes AI research agents smarter about multi-step information gathering. Instead of starting fresh each time, the system creates structured summaries after each research attempt—capturing…

Langflow, an open-source tool for building AI workflows and agents, gained attention through its GitHub repository. The platform claims to simplify deployment of AI-powered systems through a visual interface. No specific metrics, updates…

AI Adoption

OpenHands is a Python-based development tool that claims to provide AI-driven coding capabilities through autonomous agents, with integrations for ChatGPT and Claude via command-line interface. The repository positions itself as enabling…

A project called Moltbook gained attention on Hacker News after being described as "the most interesting place on the internet right now," with references to endorsement from Andrej Karpathy on Twitter. The project has rebranded to…

A developer released NanoClaw, a 500-line TypeScript alternative to Clawdbot that claims better security through Apple container isolation. Unlike OpenClaw, which runs agents with broad permissions in a single Node process, NanoClaw uses…

Cybersecurity

A developer released Zuckerman, an open-source personal AI agent that claims to modify its own code and configuration files in real time. The project aims to provide a simpler alternative to complex agent frameworks by starting with…

OpenAI has built an internal AI data agent that claims to combine GPT-5, Codex, and memory capabilities to analyze large datasets and deliver insights quickly. The company says the system can reason over massive datasets and provide…

AI Adoption · OpenAI