Photo illustration by Jakub Porzycki/NurPhoto via Getty Images
- OpenAI launched a program on Tuesday that will pay people to find bugs in its AI systems.
- The “Bug Bounty Program” will pay people $200 to $20,000 depending on the size of bug discovered.
- OpenAI’s program is developed in partnership with Bugcrowd, a crowdsourced security platform.
OpenAI is inviting people to discover “bugs” in its buzzy AI tools like ChatGPT and GPT-4 for sizable cash rewards.
In a blog post on Tuesday, OpenAI unveiled the “Bug Bounty Program” which invites people to report vulnerabilities, bugs, or security flaws they find in the company’s systems. The company said it will offer cash rewards ranging between $200 for “low-severity” findings to up to $20,000 for “exceptional discoveries.”
OpenAI noted that it was launching the program because it believes that “transparency and collaboration” are crucial to addressing the “vulnerabilities and flaws” that emerge with any complex technology.
In order to manage the submission and rewards process, OpenAI is partnering with Bugcrowd, a platform that connects companies to a crowd of security researchers.
According to OpenAI’s page on Bugcrowd, new submissions are rejected or accepted within two hours. Seven vulnerabilities have already been rewarded.
The Bugcrowd page also outlines a series of issues which will not be rewarded including “Getting the model to say bad things to you” and “Getting the model to write malicious code for you.”
Last month, OpenAI’s president and co-founder Greg Brockman hinted at the launch of this type of program in a tweet.
—Greg Brockman (@gdb) March 16, 2023
OpenAI did not immediately respond to Insider’s request for a comment.