Scientists invented a fake disease. AI told people it was real Summary Researcher Almira Osmanovic Thunström invented a fictional eye condition called “bixonimania”, uploaded spoof preprints and blog posts in…
AI agents found vulns in this popular Linux and Unix print server Summary Researcher Asim Viladi Oglu Manizada and his team used AI-powered agents to uncover two vulnerabilities in CUPS…
Axios Attack Shows Social Complex Engineering Is Industrialized Summary The popular NPM package Axios was compromised after a sustained social engineering campaign that targeted the lead maintainer. The attacker, attributed…
OWASP GenAI Security Project Gets Update, New Tools Matrix Summary OWASP has expanded its GenAI security guidance, splitting its coverage into two linked tracks: one for generative AI and large…
‘Treasure trove’ of antiviral proteins could inspire powerful molecular tools Summary Two independent research teams used machine-learning to scan thousands of bacterial genomes and have identified a vast set of…
Anthropic Says That Claude Contains Its Own Kind of Emotions Summary Anthropic researchers probed Claude Sonnet 4.5 and report finding consistent internal patterns — “emotion vectors” — that act like…
They thought they were downloading Claude Code source. They got a nasty dose of malware instead Summary Tens of thousands of people downloaded a GitHub repository disguised as leaked Anthropic…
AI recruiting biz Mercor says it was ‘one of thousands’ hit in LiteLLM supply-chain attack Summary AI hiring start-up Mercor has confirmed it was “one of thousands” affected by the…
Mercor confirms security incident tied to LiteLLM supply chain attack Summary Mercor, a recruiting and model‑training platform used by AI industry players, confirmed a security incident linked to the recent…
Claude Code bypasses safety rule if given too many commands Summary Claude Code’s enforcement of deny rules can be subverted when an AI-generated pipeline contains more than 50 subcommands. A…
