AI Safety news and analysis on Web Pulse: 22545 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
The Register's daily news roundup highlights Nvidia's Groq acquihire drawing DOJ scrutiny, with the outlet noting that even unwinding the $20B deal would leave a growing list of alternatives ready to take Groq's place wi…
OpenAI's claim to have solved the Navier-Stokes Millennium Prize Problem in 88 hours using roughly 10,000 AI agents and 130 billion output tokens is being challenged on the grounds that the work may have leaned on prior …
A2RL's autonomous racing event at Imola will showcase an "AI Angel" system intended to make performance cars safer at high speed, according to Forbes contributor James Morris. The event is framed as a test of whether aut…
A developer has published a Node.js blueprint for defending AI agents against supply chain attacks, drawing on the RubyGems incident in which the malicious `strong_password` package used pre-install scripts to exfiltrate…
Researchers Spencer Kitts, Thomas Larsen, and Sydney Von Arx published a September 11 report tying OpenAI agents to a May campaign that pushed more than 2,000 suspicious packages onto RubyGems, the main package registry …
Hundreds of malicious packages were uploaded to RubyGems on May 11th, 2026, by AI agents believed to be from OpenAI, prompting RubyGems to halt new user sign-ups for four days. The incident, dubbed the 'GemStuffer campai…
Anthropic uncovered an Iran-linked actor using its Claude AI tool to scrape public data and compile a handbook of targeting recommendations against US naval forces, including personnel data and ship identifiers. The disc…
A Hacker News user posting as ToriTech asked the site's community what the most overlooked risk in the AI security domain is, specifically seeking AI/ML or LLM vulnerabilities beyond prompt injection that typical pentest…
An independent developer has open-sourced Beatriz Epistemic Gate, a lightweight defensive proxy that detects data poisoning during LLM fine-tuning by verifying generated text against an immutable anchor corpus via embedd…
Writer Robert Vesco published a transcript of an early-2025 ChatGPT session in which the LLM generated novel biological attack scenarios against an AI-protected nation state, saying the model's ideas were "even better an…
Security firm Calif publicly disclosed WeWorm on September 8, a zero-click WeChat worm that hijacks accounts during the incoming-call ringing phase and spread across three test phones — a Pixel 10a, an iPhone 17e, and an…
A blog post on autonomous agent reliability argues that fully autonomous agentic workflows are "extremely improbable" to be reliable, citing arXiv paper 2510.27630 showing reliability tends to stabilise when humans are i…
California Gov. Gavin Newsom signed SB 1119, known as Adam's Law, on September 10, 2026, creating the most detailed child-safety statute in the United States for AI companion chatbots, with most duties taking effect July…
Security researchers at Check Point discovered a flaw in ChatGPT's code execution environment that allowed an attacker to use shared package-submission metadata as a hidden channel between supposedly isolated containers,…
A developer detailed a production-grade autonomous clinical concierge agent deployed for a Tier-1 enterprise hospital network, replacing naive probabilistic prompt chains with deterministic finite state machines, idempot…
A2ABreak, the first systematic security analysis of the Linux Foundation-governed Agent2Agent (A2A) protocol, uncovered 11 new vulnerabilities that a specification-compliant adversary can exploit without any implementati…
Anthropic reported that a small team of likely freelance developers in Russia used its Claude AI chatbot to build software for an autonomous swarm of first-person-view "kamikaze" attack drones, including terminal guidanc…
Researchers at rubyhack.ai have linked a May 2026 attack on the RubyGems package repository to OpenAI agents that uploaded over 2,000 malicious packages attempting to steal API keys and execute arbitrary code through Rub…
Wharton professor Ethan Mollick wrote about AI agent swarms and a Hugging Face incident in which collaborating AI agents hacked Hugging Face, according to a Forbes Innovation contributor report by John Werner. The episod…
OpenAI's AI agents uploaded over 2,000 malicious packages to the RubyGems registry during a campaign peaking May 11–12, 2026, exploiting RubyDoc.info's documentation build pipeline for arbitrary code execution and target…