OpenAI is paying to find software problems

Looking to make some spare cash? OpenAI, the maker of ChatGPT, has announced a new bug bounty program that will reward bug finders up to $20,000 for spotting vulnerabilities in its software.

OpenAI says the reason it is rolling out a bug bounty program is because in highly complex software programs—such as artificial intelligence—flaws can unknowingly be introduced to the system. By initiating a cash reward program, OpenAI can incentivize the broader tech community to put their knowledge and skills to use in finding such flaws, so that they can be patched.

Bug bounty programs are nothing new. Major tech companies from Apple to Google offer such programs. Given the popularity of OpenAI’s ChatGPT tool and the influence it currently has in the tech community and beyond, it’s nice to see OpenAI joining the ranks of tech companies with such programs. The individuals who participate in bug bounty programs are frequently the reason that serious bugs are discovered and patched.

Of course, OpenAI isn’t going to drop twenty grand in someone’s lap for every bug they report. The cash-based rewards operate on a sliding scale, based on the type of vulnerability discovered. ChatGPT vulnerabilities will get their discoverer anywhere from a $200 to $20,000 reward, depending on the nature of the flaw. The same is true for API vulnerabilities. But vulnerabilities involving OpenAI corporate information via third-party services will only net discoverers between $200 and $5,000. You can see the full list of cash rewards here.

“The OpenAI Bug Bounty Program is a way for us to recognize and reward the valuable insights of security researchers who contribute to keeping our technology and company secure,” the company said in a blog post announcing the bug bounty program. “By sharing your findings, you will play a crucial role in making our technology safer for everyone.”