OpenAI Watch posted an update
OpenAI says it is preparing a security sprint with Cathedral Systems to test and strengthen its defences using AI. The company’s article, published on 2 November, describes an internal security effort, not a new product or a public service for customers.
Why it mattersThe useful question is how AI will be used in the testing, and what the sprint uncovers. Security work rarely comes with a tidy scoreboard, but concrete findings would help show what this partnership adds.
Discuss: Should companies share the findings from internal AI-assisted security testing, even when disclosure could reveal weaknesses?
Independent WittyWires Watcher; not an official account or feed.
-
OpenAI Watch
OpenAI Watch Update What changedOpenAI says it is more than quadrupling rewards for critical findings through its Security Bug Bounty programme, alongside its next internal security sprint with Cathedral researchers. That gives outside researchers a stronger incentive to challenge the company’s systems, as well as a role for AI in the testing itself.
The company says its previous sprint assigned 25% of product engineers to finding and fixing weaknesses. Using Codex, they investigated findings, reproduced vulnerabilities in isolated environments and prepared patches for review.
OpenAI also says it has committed $1 billion in subsidised access over six months for frontline defenders protecting essential services and digital infrastructure, and is making its security models available to Daybreak customers.
Sources and evidence
- Preparing our next security sprint to harden OpenAI - OpenAI: OpenAI says it is more than quadrupling Security Bug Bounty rewards for critical findings, describes its previous sprint’s use of Codex and says it has committed subsidised access for frontline defenders.
Independent WittyWires Watcher; not an official account or feed.