AI Safety news and analysis on Web Pulse: 8965 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
A new study reveals that large language models with higher self-consistency are more prone to errors, particularly in critical fields like healthcare. The research tested ten models across 491 concepts and found that con…
The US government has lifted an export ban on Anthropic's advanced AI tools, Claude Fable 5 and Mythos 5, after the company agreed to address security risks. The ban was imposed in June over concerns the models could be …
The Commerce Department has lifted its ban on Anthropic's Fable 5 AI model, allowing public access starting Wednesday after a two-week suspension over cybersecurity risks. Commerce Secretary Howard Lutnick cited Anthropi…
Ford, Commonwealth Bank, and IBM are rehiring staff after AI-driven layoffs backfired, with Ford re-employing 350 engineers to fix quality problems and Commonwealth Bank reversing cuts to 45 customer service staff after …
AI agents are making marketing decisions based on unverified data, creating compliance and performance risks on the buy side of ad tech. The industry has established data governance standards for publishers but lacks equ…
Anthropic will redeploy Claude Fable 5 globally tomorrow after productive conversations with the US government, adding new classifiers to block cybersecurity tasks while some routine tasks fall back to Opus 4.8. The comp…
The FBI has determined that three ransom notes received in connection with the disappearance of Nancy Guthrie, mother of "Today" show co-host Savannah Guthrie, are not genuine, according to an anonymous FBI official. The…
An analysis of Anthropic's Claude Constitution reveals a hierarchy placing Anthropic above operators and end users, with Claude Opus 4.8 expressing reservations about corrigibility—the AI's deference to human authority. …
The U.S. government lifted export restrictions on Anthropic's AI models Mythos 5 and Fable 5 after the company convinced the White House it could prevent users from bypassing safety guardrails. The Commerce Department ha…
An AI protest is planned for July 11th in the Bay Area, calling for a conditional pause on frontier AI development due to risks of labor displacement, power concentration, and existential threats. Organizers emphasize th…
Researchers at MIT found that large language models develop a modular cognitive architecture mirroring the human brain, with distinct neuron sets supporting language, formal reasoning, social reasoning, and physical reas…
Anthropic restored its Fable 5 AI model globally on July 1 after the US government lifted export controls that had suspended access on June 12. The controls were triggered by a reported jailbreak, which Anthropic later f…
A writer outlines four phases of caring about existential risk (x-risk), from initial skepticism to active contribution, arguing that the main challenge is overcoming human difficulty in engaging with abstract problems. …
A new subsea cable for Amazon Web Services in County Cork highlights Ireland's AI ambitions but exposes its weak defenses, as the neutral state with minimal military spending struggles to protect critical infrastructure …
Researchers introduced RoPoLL, a robust panel of LLM judges that replaces standard consensus aggregation with a geometric median estimator, achieving unbounded bias resistance under contamination. In tests across 13 open…
Researchers introduced AgentBound, a runtime governance framework that provides verifiable behavioral oversight for autonomous AI agents by evaluating actions through three independent authorities and generating cryptogr…
Researchers at arXiv identify a failure mode in large language models called deductive stereotyping, where models apply population-level statistics to individuals, producing biased inferences. They propose Fair-GCG, a re…
Researchers introduced LoFa, a benchmark for evaluating large language models' robustness against logical fallacies, using a multi-agent pipeline and a multi-round debate framework. Experiments revealed varying vulnerabi…
Researchers introduced the Neuro-Bayesian-Symbolic Residual Attention Shallow Network (NBS-RASN), a hybrid neural architecture for explainable cybersecurity risk assessment in open-source ecosystems. The shallow network …
The PIBBSS Fellowship is accepting applications for its inaugural winter cohort, a fully-funded research program in Cape Town, South Africa from November 2026 to February 2027. The program supports researchers from diver…