Meta’s oversight board reviewing two cases of AI-generated nudity targeting women

Meta’s oversight board said it is inviting comments about two cases where AI-generated deepfakes targeting well-known women were posted on Instagram and Facebook

Meta’s oversight board has invited public comments regarding the rise of deepfake technology [File] | Photo Credit: REUTERS

Meta’s oversight board is inviting comments as it looks into two cases where AI-generated nude images depicting women were posted on Instagram and Facebook.

The image posted to Instagram showed an undressed Indian public figure and was posted on an account with pictures of other AI-generated Indian women. The oversight board noted that deepfakes were becoming a “problem” in the country.

The image posted to Facebook depicted an AI-generated image of a nude woman being groped, and was made to resemble an “American public figure” who was named in the caption.

While the second image had earlier been flagged as a violation and was thus easy to remove, the oversight board described a chaotic reporting process that resulted in the first image being left up on Instagram longer than needed.

(For top technology news of the day, subscribe to our tech newsletter Today’s Cache)

“In this case, a user reported the content to Meta for pornography. This report was automatically closed because it was not reviewed within 48 hours. The same user then appealed Meta’s decision to leave up the content but this was also automatically closed and so the content remained up. The user then appealed to the Board. As a result of the Board selecting this case, Meta determined that its decision to leave the content up was in error and removed the post for violating the Bullying and Harassment Community Standard,” said the oversight board.

Meta’s oversight board has invited public comments regarding the rise of deepfake technology, the way such media harms women who are public figures, Meta’s handling of deepfakes, the technology it uses for the same, and its automated reporting system.

Comments must be submitted by the end of the month.

Once the board has deliberated these cases and the submitted comments, it will share its findings, which could include policy recommendations for Meta.

While AI-powered text-to-image generators make it easy to create content featuring real people’s faces and bodies, social media tech giants have fallen behind when it comes to promptly identifying and removing such content.

Several female celebrities in India have spoken out about their shock and fear after being targeted with explicit deepfakes.

Under India’s IT Rules, social media platforms must have a grievance officer to receive complaints about morphed imagery. The social media intermediaries then have around 24 hours to remove a deepfake once it has been reported.