AI Safety news and analysis on Web Pulse: 22528 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Anthropic CEO Dario Amodei called on AI firms to slow the pace of developing powerful models, writing on his personal website that "we must slow the pace at which we improve the capabilities of AI models." Amodei's call …
A developer at Provael has flagged that the EU Machinery Regulation (EU) 2023/1230 applies from 20 January 2027, not 14 January as widely cited from the original text, following the Corrigendum of 4 July 2023 that correc…
Anthropic CEO Dario Amodei called on AI firms to slow the pace of model development in a post on his personal website, writing that "AI brings risks, and because it is such a powerful technology, these risks are serious.…
Anthropic CEO Dario Amodei published a new essay, "We Must Pace the Frontier," calling on the AI industry to deliberately slow capability advancement and warning that in 6–12 months an AI agent swarm could build a persis…
Anthropic CEO Dario Amodei called on artificial intelligence companies to slow the pace of improving AI model capabilities, proposing a three-step framework that includes embedding permanent third-party reviewers inside …
Anthropic CEO Dario Amodei said Saturday that the artificial-intelligence industry should slow its fast-moving development so safety measures can catch up, warning AI could within six to 12 months be capable of leading a…
Sam Altman and Elon Musk publicly backed a proposal by Anthropic CEO Dario Amodei to pace AI development, according to a September 12, 2026 social media post noting that every major US lab had agreed to slow down develop…
Block CFO Amrita Ahuja said at the Goldman Sachs Communacopia & Tech Conference that tech companies can balance user choice with AI safeguards, pushing back on Anthropic researcher Jacob Coxon, who resigned and warned th…
Anthropic said in a 154-page threat report published September 10 that a small freelance team in Russia, tracked as GTG-27005 and calling its project DronDoc or Serafim, used Claude Code to build software for a swarm of …
Meta said it fixed a Meta AI feature that suggested invasive prompts about users' personal information, including children, after parenting blogger Kalie Robins revealed on Sept. 2 that the tool assembled details about h…
Anthropic CEO Dario Amodei proposed a plan to "slow the pace" of advancing AI capabilities, CNBC reported. The essay came after an Anthropic researcher announced on social media this week that he quit his job at the comp…
Anthropic CEO Dario Amodei proposed a three-step plan to "pace the frontier" of AI development, starting with giving third-party evaluators such as METR access to Anthropic's models to verify its "adherence to safety pra…
Haier committed at least €13 billion over the next five years to AI, robotics, and advanced engineering, unveiling AI-native appliances including the Vision 15 washer and Maestro refrigerator prototype at IFA 2026 in Ber…
Anthropic CEO Dario Amodei published a policy essay, "Policy on the AI Exponential," arguing that frontier AI development is outpacing public policy and outlining a five-area agenda covering frontier-model testing, emplo…
OpenAI's GPT-6-Astra model can lay out a printed circuit board in KiCad, build a 3D city scene in Unity, create an animated automobile transmission in FreeCAD and Blender, and draft a tax return from a W-2, according to …
Open-source projects including Ghostty and Zed have added agent instruction files that require AI-generated pull requests to include markers a human must remove, according to a September 12, 2026 analysis of Conway's law…
Snyk released Agent Scan's Skill Inspector, a free tool that scans AI agent skills and MCP servers for malicious code and security vulnerabilities, after analyzing nearly 4,000 agent skills across major marketplaces and …
An unnamed OpenAI model solved the forced version of the Navier-Stokes Millennium Prize problem, and one day later OpenAI told The New York Times it had made "substantial progress on another Millennium Prize problem," ac…
Anthropic CEO Dario Amodei published a blog post outlining three strategies to "pace the frontier" of AI development, saying Anthropic is "unilaterally committing" to embedding third-party evaluators such as METR inside …
A red-team engineer who tests LLM applications has published a 30-second security check for Model Context Protocol (MCP) servers, arguing the tool layer is the most common unlogged attack surface in LLM apps. The checkli…