Security Negative

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

Decrypt · Sep 2, 2026

Log in to save
Informational impactNot rated
AI confidenceNot rated
Risk contextUnknownOperational, security or regulatory context
AI-ASSISTED

Feed summary

Verify with source

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.

AI output is based only on the stored feed text and may be incomplete or incorrect. It is not investment advice. Verify important claims at the original source.
Online