AI Safety news and analysis on Web Pulse: 12804 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Anthropic CEO Dario Amodei and Google DeepMind CEO Demis Hassabis proposed a US-led AI coalition at the G7 summit, with Canadian Prime Minister Mark Carney agreeing the US could lead. The closed-door lunch in Évian-les-B…
A blog owner is blocking old browser user agents to combat high-volume crawlers, including those used for LLM training, causing legitimate users of outdated browsers to be denied access. The block also affects feed reade…
Western leaders and top AI CEOs agreed at a G7 meeting in France to collaborate on regulating advanced AI models to counter China's influence, despite recent U.S. restrictions on Anthropic's models causing tensions among…
A technologist expresses internal conflict over AI, agreeing with accelerationists, artists, nationalists, doomers, and openness advocates simultaneously, highlighting the contradictions in current AI debates. The author…
RootSign, a new SDK that produces cryptographic audit logs for CrewAI and LangGraph agents, was released on GitHub. The tool adds a SHA-256 hash chain to every tool call to detect log tampering, addressing the lack of le…
Anthropic CEO Dario Amodei urged G7 leaders to maintain unified AI regulation at a summit lunch with President Trump, days after the Trump administration imposed export controls on Anthropic's 'Fable' AI model over cyber…
Port has identified seven categories of hidden infrastructure debt that arise when AI agents move from local demos to production in enterprise systems, including integrations, observability, governance, human-in-the-loop…
The White House ordered Anthropic to block all jailbreaks of its Claude Fable 5 AI model after the Commerce Department used Export Administration Regulations to force the model offline on June 12, citing cybersecurity ri…
Smart fridges and other internet-connected appliances risk losing functionality and security when manufacturers stop providing software updates, a process known as 'zombification.' A 2025 FTC report found that 89% of sma…
The Trump administration ordered Anthropic to block access to its Fable 5 and Mythos 5 AI models for all foreign nationals, including those in the US and the company's own employees, citing national security authorities.…
Britain lobbied the White House for an exemption from the Anthropic AI export ban but was told there was "zero chance." The rejection exposes the UK's dependence on American AI and strengthens the case for sovereign alte…
Illinois could become the first U.S. state to make driving while wearing AI smart glasses a felony, after the General Assembly approved a bill updating distracted-driving law with no hands-free or stopped-traffic excepti…
AI researcher Gary Marcus warns that a weak OpenAI IPO could trigger a cascade of financial problems for Nvidia, Oracle, and CoreWeave, as their valuations depend on OpenAI's massive chip and data center demand. OpenAI i…
Los Alamos National Laboratory has developed a method to detect hallucinations in vision-language AI models, improving reliability in critical applications. The technique identifies when AI generates incorrect or fabrica…
A developer ported the MACHIAVELLI benchmark, which measures unethical AI agent behavior, to the Inspect evaluation framework to make it easier for evaluators to use. The re-implementation is now officially part of Inspe…
Only 16% of Americans believe AI will benefit society over the next two decades, according to Pew Research Center, even as 21% of U.S. workers use AI chatbots in 2025. The public's top fears include job elimination, eros…
Anthropic CEO Dario Amodei and DeepMind CEO Demis Hassabis called for a U.S.-led AI coalition during a closed-door lunch at the G7 summit in Évian-les-Bains, France, with around a dozen tech executives including OpenAI C…
Three aircraft crashes in Missouri, California, and Texas killed 21 people in recent days, prompting federal investigations by the National Transportation Safety Board. The accidents involved a business jet in Laredo, Te…
Consulting firm KPMG withdrew a report on Agentic AI after AI detection company GPTZero found it contained AI-generated fabrications and mangled citations. The report, titled 'Total Experience: Redefining Excellence in t…
Anthropic CEO Dario Amodei said he left OpenAI because he didn't trust Sam Altman, stating 'why argue with someone when you don't trust them?' in a Bloomberg interview. Amodei expressed peace with the rivalry, saying the…