Claims: OpenAI's Astra Becomes Its First AI Model With 'Critical' Hacking Abilities

These are attributed source records, not independent verification of each assertion.

Complete claim record (JSON) · Original source

1. title: OpenAI's Astra Becomes Its First AI Model With 'Critical' Hacking Abilities (alleges; supported) Source
2. url: https://decrypt.co/377180/openai-astra-first-ai-model-critical-hacking (alleges; supported) Source
3. summary: 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. (alleges; supported) Source
4. published: Wed, 02 Sep 2026 16:43:53 +0000 (alleges; supported) Source