Twitter has launched its first ever artificial intelligence bug bounty program, challenging people to find bias in its image-cropping algorithm and offering cash prizes for those who best assess the models.
The winners — chosen based on a rubric created by Twitter's Machine Learning Ethics, Transparency and Accountability or META team — will be announced at this year's DEF CON AI Village, and the first place prize will be $3,500.
"We're inspired by how the research and hacker communities helped the security field establish best practices for identifying and mitigating vulnerabilities in order to protect the public. We want to cultivate a similar community, focused on ML ethics, to help us identify a broader range of issues than we would be able to on our own," Rumman Chowdhury and Jutta Williams, META's leader and product manager respectively, wrote in a company blog post.
Twitter removed the image-cropping algorithm from public use after small instances of bias were uncovered in research by the META team earlier this year. The image-cropping assessment was the first public change created by the new team after Chowdhury, a noted AI ethicist, joined the company to lead its ML transparency work alongside Williams and Ari Font, the engineering director of Twitter's ML platform.