OpenAI raises critical bug bounty rewards to more than four times their previous level

發生什麼事

@OpenAI says its next in-house security exercise will involve Cathedral Systems. Cathedral researchers and OpenAI staff will use OpenAI’s models to probe the company’s defenses, then find and repair weak points. As AI gets better at finding vulnerabilities, defenders need to adopt it before attackers follow suit.

The public bug bounty program is also changing: awards for findings rated critical will rise to more than four times their previous level. OpenAI says it will keep advancing its models, apply them to strengthen its own systems, and make them available to more trusted security defenders.