AI Safety news and analysis on Web Pulse: 12802 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
A new report examines the failure of Trade Adjustment Assistance, established over 60 years ago to help workers displaced by globalization, and draws lessons for policymakers preparing for potential job losses from artif…
Google's Pixel Screenshots app is expanding beyond on-device AI to include cloud processing, using the company's Private AI Compute technology to keep data secure. The change, rolling out in version 1.26.134.11, updates …
Researchers at Aikido Security discovered 15 malicious plugins on the JetBrains Marketplace that stole AI API keys from approximately 70,000 developers. The plugins, posing as AI coding assistants and Git utilities, secr…
The Trump administration imposed export controls on Anthropic's Fable 5 and Mythos 5 AI models, citing national security, prompting Anthropic to disable the models for all users including in the US. The move overshadowed…
Multi-agent AI systems in production are failing due to distributed systems problems like deadlocks, not AI model quality, according to industry reports. LangChain's 2026 State of Agent Engineering survey found 57% of or…
NewEdge Advisors has deployed Anthropic's Claude directly to its field advisor teams for portfolio analysis, meeting preparation, and client prospecting, with compliance guardrails and data-security controls. The firm, p…
Payment fraud costs businesses over $100 million annually, driving banks, merchants, and fintechs to deploy machine learning systems that analyze hundreds of data points in milliseconds to block unauthorized transactions…
China's top diplomat Wang Yi called for a global AI cooperation organization offering cheap or free models to developing nations, as the G7 discussed limiting access to US AI models to "trusted partners." The competing v…
A security researcher known as Bobdahacker discovered a critical authentication flaw in FIFA World Cup systems, allowing unauthorized access and content injection. Google Chrome's Manifest V3 update will significantly re…
OpenAI has joined a coalition allowing the US government pre-release access to its most advanced AI models for safety testing, part of a broader industry shift toward voluntary cooperation with federal regulators. The mo…
The Internet Engineering Task Force (IETF) is considering standards changes that would allow websites to block automated access, threatening the open web. Proposals from the AI Preferences and Web Bot Auth working groups…
Anthropic CEO Dario Amodei published an essay urging democratic nations to form a coalition for binding AI safety regulations, including mandatory third-party audits and government authority to halt dangerous AI deployme…
The Cybersecurity and Infrastructure Security Agency has received full access to Anthropic's Mythos Preview model, a powerful AI tool for vulnerability scanning, but the White House has not yet set clear usage parameters…
French President Emmanuel Macron pushed at the G7 summit to restore allied access to Anthropic's Mythos AI, a frontier model that the US ordered restricted due to its ability to find zero-day vulnerabilities. Macron expe…
President Trump signed an executive order on June 2 establishing a voluntary pre-release review framework for frontier AI models, asking developers to give the government access up to 30 days before public release for cy…
The G7 concluded a summit in Evian-les-Bains, France, with leaders discussing artificial intelligence and a surprise US-Iran agreement. French President Emmanuel Macron praised the Iran deal as a "very good deal" and off…
The Trump administration restricted use of Anthropic's Fable 5 and Mythos 5 models under export-control regulations, forcing the company to suspend access within 90 minutes. Defense contractors like Lockheed Martin remov…
Government-associated AI policy leaders met at the British High Commission in Ottawa to discuss forming a coalition of "middle powers" to create AI safety guardrails distinct from the United States and China, which parti…
Canadian organizations temporarily lost access to Anthropic's AI models during an outage, highlighting risks of relying on foreign AI providers. Experts say the incident underscores Canada's underinvestment in domestic A…
A study by RMIT University surveying over 16,000 adults across ten countries found that 14.5% have experienced sextortion, with 4.8% admitting to perpetrating it. The Australian eSafety Commissioner has launched an aware…