On Friday, Twitter Inc. said it would launch a competition for computer researchers and hackers to identify biases in its image cropping algorithms, after a group of previous researchers found that the algorithms tended to exclude blacks and men.
The competition is part of a broader effort across the tech industry to ensure AI technologies operate ethically.
The social networking company said in a blog post that the bounty contest aims to identify the “potential dangers of this algorithm beyond what we have identified ourselves.”
After criticism last year about previewing images in posts that didn’t include black faces, the company said in May that a study by three machine learning researchers found an 8% difference in demographic parity in favor of women, and 4% in favor of white individuals.
Twitter publicly released computer code that decides how to crop images in its Twitter feed, and on Friday said participants were asked to see how algorithms could cause harm, such as profiling or demeaning groups of people.
Winners will receive cash prizes ranging from $500 to $3,500 and will be invited to present their work at a workshop hosted by Twitter at DEF CON in August, one of the largest hacker conferences held annually in Las Vegas.
–
–
“Hardcore pop culture expert. Game player. Internet fan. Trouble maker. TV fan. Loyal social media fan.”
—
–
