OpenAI launches bug bounty program with rewards up to $20,000

OpenAI just invited the world's most talented hackers to break into its systems, and it is ready to pay up for the privilege.

The AI lab officially launched its Bug Bounty Program, calling on security researchers, ethical hackers, and tech enthusiasts worldwide to hunt down vulnerabilities across its platform.

To keep things structured, OpenAI partnered with leading vulnerability platform Bugcrowd to manage disclosures and payouts.

Big cash for big bugs

Researchers will be rewarded based on the severity and impact of the flaws they uncover.

Payouts start at $200 for low-severity issues and climb up to $20,000 for exceptional, critical discoveries.

Alongside the cash rewards, OpenAI used the announcement to plug open full-time positions across its internal security team.

Why it matters

OpenAI is growing up fast. Moving from an experimental research lab to an infrastructure backbone serving hundreds of millions of users means security cannot just be an afterthought handled behind closed doors.

Partnering with Bugcrowd creates a clear, legal lane for white-hat researchers to report flaws without the fear of legal pushback, while incentivizing them to disclose vulnerabilities before bad actors exploit them.

However, traditional bug bounties are historically built for standard web applications, authentication glitches, and infrastructure leaks. It remains to be seen how well these programs will handle frontier AI risks like novel jailbreaks and prompt injections. Still, offering cold hard cash is the smartest first step to securing a rapidly expanding attack surface.