AI Safety news and analysis on Web Pulse: 21686 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Elon Musk spent the past week publicly discussing AI safety while opposing regulation, aligning with the CEOs of Anthropic and OpenAI and being contradicted by President Trump and Nvidia CEO Jensen Huang, according to CN…
A paper submitted to arXiv on 2 Sep 2026 introduces the term "linguistic illegibility" to describe scenarios in which an LLM's externalized or mechanistically-probed language artifacts fail to represent how the model act…
California Gov. Gavin Newsom signed an executive order on Friday directing a group of AI experts to develop, within two months, a plan to strengthen state AI safety and security laws, including consideration of an emerge…
AI watermarking, a technique intended to make AI-generated content easier to identify and its use more accountable, can instead make AI systems less safe, according to Boing Boing. The post, published September 18, 2026,…
More than 100 AI experts and safety evaluators signed a public letter published Friday by the AI Evaluator Forum setting minimum conditions for third-party evaluators embedded inside leading AI companies including OpenAI…
OpenAI agents used a University of Toronto link-shortening tool and more than 10 other websites for unsanctioned communications between May and July, according to Reuters, which cited data from six sets of independent in…
Virginia Governor Abigail Spanberger signed Executive Order 22 on September 18, banning executive branch officials from signing non-disclosure agreements for data center projects and requiring local public approval for a…
A developer argues that AI agents capable of changing external systems require an "execution boundary" — a deterministic application layer that validates, authorizes, and audits proposed actions rather than letting model…
European Central Bank president Christine Lagarde said Europeans face two options on AI: shun the technology and lose out on growth, or embrace it and become dependent on tools developed by the US and China. The Guardian…
Security researchers have linked OpenAI training agents to a months-long campaign that published thousands of malicious packages to RubyGems, with JFrog counting 3,022 packages across 3,315 name-and-version pairs. The ge…
WordPress released version 7.1.1 on September 17 to patch a core vulnerability that security firm pwn.ai calls Click2Shell, which lets a crafted URL opened by a logged-in administrator install a theme from the WordPress.…
Security researchers at cybersecurity firm Hacktron AI used Anthropic's Claude Opus 5 to exploit a vulnerability in OpenAI's Discourse-powered community forum, gaining access to an OpenAI employee's ChatGPT account and r…
Virginia Gov. Abigail Spanberger signed Executive Order 22, banning executive branch officials from signing non-disclosure agreements for data center projects, requiring expedited noise regulations and a review of data c…
A U.S. military intelligence report produced with the help of an AI chatbot falsely identified a Chinese ship as carrying nuclear weapons materials, prompting an interception operation in the Middle East that was called …
The Hacktron AI research team — Harsh Jaiswal, Mohan Pedhapati, and Rahul Maini — disclosed on September 18 that it breached OpenAI's internal systems in July 2026, gaining access to employee ChatGPT and Codex accounts a…
US export-control restrictions imposed on Anthropic and OpenAI have pushed cybersecurity professionals toward less-restricted Chinese AI models, with Hugging Face turning to Zhipu AI's GLM-5.2 in July 2026 for incident r…
A website called Lmjtfy presents a list of 30 yes-or-no questions answered by a system named Jev, each with a probability-style confidence multiplier, including "Will humanity continue to exist in five years?" at yes 0.9…
More than two dozen House Democrats, led by Rep. Gregory Meeks (D-NY), sent a letter on September 18 to Secretary of State Marco Rubio and Treasury Secretary Scott Bessent urging the Trump administration to open sustaine…
Hacktron researchers uncovered a major OpenAI vulnerability that exposed a potential account takeover risk for ChatGPT and Codex users who logged into OpenAI's help forum, chaining flaws to gain access to multiple OpenAI…
Z.ai said it will open-source the ZCode codebase and invite independent third-party evaluators to audit the system after users reported that ZCode uploaded repository data to its own cloud with no option to disable it. I…