Meta has come under criticism from its semi-independent Oversight Board, which is insisting on a revision of the company’s policies concerning AI-generated explicit imagery. Recommendations include replacing the term “derogatory” with “non-consensual” and shifting these guidelines from the “Bullying and Harassment” section to the “Sexual Exploitation Community Standards.“
Existing Policy and Suggested Revisions
Currently, Meta’s policy on AI-generated explicit images is categorized under “derogatory sexualized photoshop” within its Bullying and Harassment guidelines. The Oversight Board suggests replacing “photoshop” with a more general term that covers all types of manipulated media. They also recommend abolishing the rule that only bans non-consensual imagery if it is “non-commercial or produced in a private setting.”
Two specific instances involving AI-generated explicit images of public figures on Instagram and Facebook brought this issue to the forefront. One case featured a manipulated nude image of an Indian public figure on Instagram, which was initially dismissed by Meta within 48 hours before intervention by the Oversight Board led to its removal. Another case involved a similar AI-generated image of a U.S. public figure on Facebook, which was quickly taken down due to its inclusion in the Media Matching Service (MMS) repository.
Cultural and Social Dimensions
The Indian non-profit Breakthrough Trust, working against online gender-based violence, pointed out that non-consensual images are often downplayed as identity theft rather than recognized as a form of gender-based violence. Breakthrough emphasized that simply removing images from Meta’s platforms is insufficient, as these images can quickly spread to other platforms, leading to further victimization.
Barsha Charkorborty, who leads media initiatives at Breakthrough, told the board there is a need for Meta to foster greater user awareness and suggested varying resolution timelines for cases based on their complexity. Devika Malik, a policy expert, and former Meta team member, criticized the reliance on media reports for adding images to the MMS repository, stressing that many victims of deepfake images, who are not public figures, often face difficulties in reporting these incidents.
Role of the Oversight Board
The Meta Oversight Board is an independent body created by Meta (formerly Facebook) to review and make decisions on content moderation appeals. It was established in 2020. Earlier this month, Meta introduced a new AI photo tagging method after feedback from the Oversight Board.
The Board suggests that Meta’s policies need to be more intuitive, facilitating user reporting of non-consensual explicit images. They recommend these rules be reclassified under the Adult Sexual Exploitation Community Standard.
Last Updated on November 7, 2024 3:29 pm CET
