image credit: wikipedia .Meta AI logo is seen in this illustration taken September 28, 2023

The Oversight Board reported Thursday that Meta failed to remove an explicit, AI-generated image of an Indian public figure until it was questioned by the board, calling on the tech giant to do more to address non-consensual, nude deepfakes on its platforms.

In April, the Meta Oversight Board launched a probe into Meta’s handling of deepfake pornography, including two incidents where explicit photos of American and Indian prominent figures were posted.

Online harassment of celebrities like Taylor Swift, US high school students, and other women has raised awareness of AI-generated pornography. Easily available generative AI techniques make creating such graphics faster, easier, and cheaper. Social media networks like Meta, where such photos propagate quickly, are under pressure to address the issue.

Facebook immediately removed the artificial intelligence-generated image of the American public figure, which showed her nude and being groped. The image had previously been added to a matching bank that automatically detects rule-breaking images. Although Meta received two reports on the Indian public figure, the business did not delete the image from Instagram until the Oversight Board intervened.

The Oversight Board reported that Meta withdrew the post for breaking the bullying and harassment community standard after realizing its mistake. Meta disabled the account that posted the information after the board began deliberations.

The analysis implies that Meta is not regularly enforcing its rules against non-consensual sexual imagery, even as AI makes this sort of harassment more widespread. Meta has been criticized for its content moderation in non-Western or non-English-speaking nations, according to the research.

Meta said it supported the board’s decision. Along with removing the specific posts mentioned in the report, the company will “take action” on pictures of the Indian public figure that are “identical and in the same context” as those the Oversight Board highlighted “where technically and operationally possible to do so.”

The quasi-independent Oversight Board, composed of specialists on freedom of expression and human rights, issued more recommendations for Meta to combat sexualized deepfakes in its report. It encouraged the business to clarify its ban on “derogatory sexualized Photoshop” by including “non-consensual” and including AI and other photo editing methods.

The report stated that Meta informed the board that it had not initially included the Indian public figure’s image in its rule-violating photo matching bank because there had been no news reporting about it, unlike the US public figure’s. “This is worrying because many victims of deepfake intimate images are not in the public eye and are forced to either accept the spread of their non-consensual depictions or search for and report every instance,” the Board said. Meta could also consider whether an image was AI-generated before adding it to the bank.

Meta’s fight against non-consensual deepfakes is part of their strategy to avoid user sexual exploitation. The business announced Wednesday it had erased 63,000 Nigerian financial sextortion accounts, where teens are duped into sharing nude photographs and extorted.