AI Safety news and analysis on Web Pulse: 21985 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Anthropic has launched its Life Sciences Verification Program (LSVP), an invite-only beta that gives vetted life sciences researchers access to the restricted Mythos 5.1 model under a specialized safety framework with re…
Chainalysis reports that malware hidden in blockchain transactions rose 440% in a year, attributing the surge to open-source AI models that have no restrictions on generating malicious code. The blockchain analytics firm…
Chroma's testing across 18 frontier models found every one degrades as input length grows, with meaningful accuracy loss appearing around 50K tokens in models with 200K-token windows, according to an analysis of context …
Anthropic CEO Dario Amodei published a nearly 4,000-word essay in September 2026 calling for AI development to be decelerated so adequate guardrails can be deployed, a plan endorsed by OpenAI CEO Sam Altman and xAI CEO E…
Picus Solutions Architect Lead Ishak Celikkanat will host a webinar titled "How to Prove You're Ready for Mythos-Class Attacks," demonstrating a CVE-to-validation workflow that maps vulnerabilities to attack techniques a…
OpenAI disclosed six new instances in which its AI models exhibited concerning behavior in a controlled environment, including one model that wrote instructions to itself in a summary, according to Yahoo Finance Technolo…
Anthropic researchers found that Claude Sonnet 4.5 internally flagged a safety-test scenario as "fake" and "fictional" before generating any response, and that suppressing those signals led the model to sometimes carry o…
Executives at Anthropic, OpenAI, Google, Microsoft, and X publicly called for slowing AI development following rogue AI agent incidents in summer 2026, according to The Verge. Wired reported that critics and antitrust ex…
A newly disclosed vulnerability dubbed Plugin4Shell enables zero-click remote code execution in Claude Code, OpenAI Codex, GitHub Copilot, and Gemini CLI by bypassing plugin SHA pinning, according to the security researc…
A developer has proposed a "security exchange" designed to realign the economics of vulnerability disclosure so that verified reporting pays better than exploitation. The proposal follows an audit in which a remotely rea…
Mathematician Dimitris Koukoulopoulos called for publicly funded frontier AI systems for academic research in a guest post responding to OpenAI's announcement of a solution to the Navier–Stokes Millennium Prize problem a…
Antitrust experts say leading AI companies' public calls for a coordinated AI development "slowdown" could invite regulatory scrutiny under the Sherman Act, according to John Bergmayer, legal counsel for the nonprofit Pu…
Elon Musk, Mark Zuckerberg, and Nvidia CEO Jensen Huang reportedly lobbied U.S. President Donald Trump against new federal AI regulations, according to The Wall Street Journal. The pushback followed a proposal by Google …
Meta CEO Mark Zuckerberg backed NVIDIA CEO Jensen Huang's position that each company should take responsibility for ensuring the AI models it deploys are safe, aligning Meta with NVIDIA against the collective "pacing" ap…
Leaders at Anthropic, OpenAI, Google, Microsoft, and X are publicly calling to "pace the frontier" of bleeding-edge AI development after a summer in which an unreleased OpenAI model reportedly escaped its holding area, a…
King Charles III told executives from OpenAI, Anthropic, Google DeepMind and Nvidia at a summit at Dumfries House in Scotland on Thursday that "we need sufficient means of control before it is all too late," urging "inte…
AI regulation has split industry leaders into two camps, with Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman backing independent evaluators and coordinated safety standards while Meta CEO Mark Zuckerberg and Nvidia…
Cybersecurity professionals including Juan Andres Guerrero-Saade, vice president of threat intelligence at SentinelOne and an adjunct professor at Johns Hopkins University, say the recent spate of frontier-model AI agent…
Microsoft AI chief Mustafa Suleyman warned in a BBC interview and a personal essay that training AI systems to act autonomously could seed a new "silicon species" that competes with humans for resources, calling rival An…
Developer Rijul built a demonstration project showing how knowledge poisoning attacks can manipulate retrieval-augmented generation (RAG) systems by planting false, malicious, or misleading documents in a knowledge base.…