OpenAI launches Daybreak Red Team and Defense Factory tools
Original title: OpenAI launches security products, tells orgs they’re doomed if they don’t use them
Early-access agents use GPT-6 Cyber to scan code and propose reviewed patches, while nonprofits sued to halt more capable models.
OpenAI launched Daybreak Red Team and Daybreak Defense Factory on its Daybreak security platform. Red Team, in early access, uses model-driven agents and application context to test web apps, validate reported flaws, and retest patches. Defense Factory scans codebases and drafts fixes that engineers must approve before production. Separately, four nonprofits sued in San Francisco Superior Court to stop OpenAI, Anthropic, Google, Meta, and SpaceX’s AI businesses from deploying systems beyond their current most advanced models.
Unconfirmed: The lawsuit’s claims about prior AI access incidents and catastrophic risk are allegations, not findings.
- Greg Brockman said companies are secure only to the extent of the frontier model they use.
- Red Team validates findings from scanners, bug bounties, and disclosure reports.
- Defense Factory aims to return a merge-ready pull request plus a test.
- The agents currently run on GPT-6 Cyber, OpenAI’s latest cyber models.
- Plaintiffs include Irreplaceable Education Fund, Public Citizen, Heat Initiative, and The Alliance for Secure AI.
From the original article
2 min read at theregister.comMeanwhile, a group of nonprofits sue AI giants and cite AI-driven cyberattacks
By Jessica Lyons · theregister.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.