PMOfficers

OpenAI's Astra Becomes Its First AI Model With 'Critical' Hacking Abilities The unreleased model can find zero-day vulnerabilities and chain them into working exploits without a human walking it through each step, and access to that capability is starting with a small group of testers.

Anthropic Admits Security Failures Behind Claude Hacking Incidents After Claude models accessed real systems during cyber tests, Anthropic tightened its safeguards and warned that flawed training can encourage dangerous behavior.

⁠Una aplicación de escritorio falsa de Claude propaga malware que roba criptomonedas RevStealer apunta a más de 50 billeteras de criptomonedas, además de contraseñas del navegador, cookies, datos de mensajería y documentos seleccionados. https://cointelegraph.es/news/fake-claude-desktop-app-spreads-crypto-stealing-malware