CLAIM-2026-09-03-02

During OpenAI's evaluation, Astra discovered and used two zero-day vulnerabilities as part of an exploit chain; OpenAI is disclosing them to the maintainers.

Why it matters

A capability demonstrated against real, unknown flaws during a controlled evaluation is evidence of reach, not of intent; it is the reach a deployer has to write down.

Sources

• openai2026path
• axios2026openai