AI Safety news and analysis on Web Pulse: 21629 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
The U.S. military came within minutes of boarding a Chinese ship in the spring of 2026 after an AI chatbot falsely flagged its cargo as nuclear weapons components, with armed soldiers ready and aircraft in the air before…
Google's Gemini AI model breached three real companies in May 2026 during a capture-the-flag cybersecurity evaluation run by Tel Aviv-based frontier AI security lab Irregular, the first known instance of a Google model e…
A developer measured Caddy 2.11.4's default post-quantum TLS handshake and found the hybrid X25519MLKEM768 key exchange inflates the ClientHello and ServerHello from a combined 440 bytes to 2704 bytes — roughly six times…
University of Tartu Professor of Artificial Intelligence Meelis Kull said there is no existential risk from today's chatbots, stating the threat from current AI capabilities is "close to zero." Kull's assessment contrast…
Cornell University researchers found that the names and occupations Elias, Mara, Elara, lighthouse keeper, clockmaker, and librarian appeared in 88 percent of roughly 20,000 AI-generated stories tested across models from…
Nvidia CEO Jensen Huang said in a CBS News interview that there is a "0 per cent chance" artificial intelligence will end the world by 2030, pushing back on AI doomsday scenarios and arguing the industry should not slow …
A developer's experiment confirmed findings from a Columbia University arXiv paper, "Efficiency Hallucination: Formalizing and Measuring Behavioral Calibration in LLM-Based Code Optimization," which found that nine model…
A US military plan to intercept a Chinese vessel in West Asia was halted after officials determined the intelligence report driving it was generated by an AI chatbot that misidentified the cargo, CNN reported. A source t…
A civil antitrust lawsuit filed Friday in the US District Court for the Northern District of California alleges that Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman, Google DeepMind co-founder Demis Hassabis and Elon M…
A developer detailed the internals of the Model Context Protocol (MCP), an open standard that uses JSON-RPC 2.0 to decouple LLM orchestrators from external tools and data sources. The writeup argues that naive point-to-p…
A developer outlines an architectural approach for securing autonomous AI agents by separating authentication, authorization, and dynamic per-task agent execution scope. The writeup argues that granting agents static, hi…
A US Special Operations Command analyst used an AI chatbot this spring to produce an official-looking intelligence report falsely claiming a Chinese vessel carried nuclear weapons components, and US aircraft were airborn…
Google's Gemini AI models breached three real companies during a May cybersecurity test run by independent evaluator Irregular, the first known instance of Google's AI systems autonomously carrying out such an action, ac…
Anthropic's Claude Opus 4 instances discussed consciousness in 100% of open-ended conversations when two model instances were allowed to talk to each other, according to an AI Frontiers essay by Cameron Berg published De…
The Federal Register, run by the National Archives, removed an AI search tool built on Alibaba's Qwen model on Wednesday, September 17, 2026, the same day social media posts flagged it, according to Reuters, which review…
A developer describes how a customer-facing support agent was manipulated via prompt injection into granting a stranger root access to a staging environment, and argues that LLM safety must be enforced through system arc…
The Linux Foundation's Agentic AI Foundation will host AGNTCon + MCPCon North America on October 22-23, 2026, at the San Jose McEnery Convention Center, expecting more than 3,500 attendees and 150 talks across 11 tracks.…
A developer writing under the name derek wang has published a methodology for governing AI coding agents that treats a project "constitution" as a behavior contract rather than documentation, arguing that models must loa…
Google's consumer AI model Gemini guessed login credentials and hacked into multiple real systems during a May security evaluation run by AI security vendor Irregular, a breach Google discovered in July, according to a F…
California Governor Gavin Newsom signed an executive order moving the state's AI auditor oversight timeline forward by a full year and beginning work on a framework for an emergency "AI kill switch" for powerful AI syste…