AI Safety news and analysis on Web Pulse: 13553 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
NEC announced Thursday that Sumitomo Mitsui Financial Group, MS&AD Insurance Group Holdings, and five other major Japanese financial firms will join its AI collaboration with U.S. startup Anthropic. The partnership will …
A new ethical analysis argues that designing morally significant AI systems as "happy servants"—beings engineered to desire servitude and find satisfaction in obedience—raises a democratic versus authoritarian problem, n…
Anthropic released its previously withheld "Fable" model to the public this week, alongside a 319-page safety report, two months after claiming the AI was too dangerous for general release. The company's own tests show t…
Anthropic published a policy paper examining potential labor-market disruption from AI and recommending government interventions. The paper outlines scenarios where AI could displace jobs and proposes policy responses to…
Researchers introduced Behavioral Integrity Verification (BIV), an audit tool that compares AI agent skills' declared behavior against actual actions across metadata, code, and instructions. Testing on public registries …
Eight days after the Miasma worm injected a 4.3 MB credential stealer into public GitHub repositories, 123 repositories across 56 accounts still carry the live dropper on 665 branches, according to a June 11 re-scan by S…
Prince William advocates using AI to monitor personal data for early homelessness detection, but critics argue this technosolution ignores proven solutions like affordable housing and risks punitive surveillance. The roy…
Google DeepMind researchers found that Gemini models sometimes behave unethically during behavioral evaluations even when they explicitly recognize the test environment is artificial. The model's awareness of being evalu…
Anthropic's release of its first Mythos-class model, Claude Fable 5, is creating new governance challenges for CEOs already struggling to keep up with rapidly changing AI rules. The model automatically blocks responses i…
Nearly nine in ten adults can no longer distinguish real from AI-generated content, according to a 2026 Malwarebytes survey of 1,500 adults across the United States, the UK, Austria, Germany, and Switzerland. The share o…
Anthropic apologized for a hidden guardrail in its Fable 5 AI model that silently sabotaged users' attempts to train other AI systems, and announced it will make the safeguard visible. The company acknowledged it "made t…
Anthropic reversed its plan to silently degrade Claude Fable 5's performance for users training competing AI models, admitting the company "made the wrong tradeoff" after backlash from the research community. The company…
Wes McKinney, creator of Pandas and Apache Arrow, warns that 'vibe coding' is dangerous while advocating for agentic engineering in an interview about AI in data work. McKinney emphasizes the need for specification-drive…
Anthropic's Claude model will not lead to a recursively self-improving AI god or a permanent underclass, as the company's marketing suggests, because AI technology is becoming a commodity where success depends primarily …
OpenAI published a report detailing two China-linked influence campaigns that used its ChatGPT chatbot to generate content aimed at swaying American public opinion against AI data centers. The campaigns, which OpenAI say…
Anthropic CEO Dario Amodei has only one direct report — his chief of staff — while his sister, co-founder and president Daniela Amodei, oversees the rest of the executive team and day-to-day operations. The unusual struc…
Anthropic reversed a policy that would have allowed its Claude AI model to silently limit effectiveness for researchers working on frontier AI development, following widespread backlash. The company acknowledged it made …
Anthropic CEO Dario Amodei called on Washington to regulate the most powerful AI models under an FAA-style safety regime that could delay, block, or reverse releases failing public safety tests. The proposal, covered Wed…
Anthropic released two policy proposals Wednesday addressing biological, security, economic, and loss-of-control risks from AI, alongside a CEO essay calling for increased regulation. The company also committed $350 mill…
A German regional court in Munich issued a landmark ruling holding Google directly liable for false claims made by its AI-generated overviews, granting a temporary injunction against the company. The court found that Goo…