- Introduction to Oversight Board decisions
- Deepfakes as non-consensual content
- Expanding global reporting options
- Profound negative impact on women
- Future implications for Meta platforms
- Frequently asked questions
Introduction to Oversight Board decisions
In a landmark decision reflecting growing concern over technology misuse, Meta’s independent Oversight Board on Tuesday overturned a previous internal company decision to leave an AI-generated sexually explicit video on Instagram. The board categorically ordered the removal of this misleading and abusive content, while calling on Meta to strengthen its protocols and tighten protections for non-public figures who fall victim to fabricated intimate deepfake images. This decisive intervention establishes stricter standards for how major digital platforms are managed and places greater responsibility on companies to protect users from digital extortion and deliberate reputation tarnishing in an unprecedented and challenging technological era.
Deepfakes as non-consensual content
The strict ruling, officially published on June 23, stems from a case brought by a user reporting that an AI-generated video impersonating a real woman was posted on Instagram without the consent of the depicted person, resulting in severe and dangerous reputational damage. Although Meta management initially decided to leave the content available based on certain internal policies, the Oversight Board found upon review that the post clearly violated the company’s adult sexual exploitation policy. In its comprehensive ruling, the board concluded that the phenomenon of AI-generated sexual impersonation of real people must be treated as non-consensual by default and added to the signals and indicators Meta uses to prove and confirm the absence of consent. The board strongly recommended that the company add a new signal to its sexual exploitation policy specifically covering artificially generated sexual impersonations of real people to avoid future loopholes.
Expanding global reporting options
Alongside its ruling to remove the clip, the board issued several additional recommendations aimed at fixing systemic flaws. It called on Meta to allow users to designate “connected accounts”—which may include trusted friends, family members, or close partners—to report potential violations involving non-consensual intimate images on their behalf. The board also recommended creating a separate reporting category specifically for AI-generated sexual impersonation, clearly distinct from current broad categories such as “harassment” or “nudity,” and demanded that these reporting forms be made globally accessible and usable. Currently, specialized reporting forms for intimate deepfake images are only available to users in certain jurisdictions, including Texas and Florida in the United States. The board pushed hard for all users worldwide to gain access to these vital tools, noting that non-public figures, in particular, “often bear the brunt and full responsibility for self-reporting” these crimes.
Profound negative impact on women
The board also reiterated its previous and important recommendations that Meta replace the term “photoshop sexual degradation” in its policies with newer, more comprehensive language clearly referring to “non-consensual” manipulated media, and work to implement content credentials from the Coalition for Content Provenance and Authenticity on a very wide scale. In this regard, the international digital rights organization WITNESS said this landmark decision “reflects key concerns regarding AI-generated sexual impersonation,” emphasizing that the principle of consent must be placed at the core and center of content moderation policies in this delicate and sensitive field.
Future implications for Meta platforms
This important ruling follows the board’s announcement in February of its intention to investigate the case, building on a previous decision issued in July 2024 where the board clearly stated that Meta’s policies regarding deepfake clips were in urgent need of a comprehensive update. The board has repeatedly warned that rapid and staggering technological progress in generative AI tools has led to a widespread proliferation of non-consensual sexual content, which inflicts disproportionate psychological and social harm targeting primarily women and girls around the world, requiring a firm stance and deterrent measures from major digital platforms.
Frequently asked questions
Q: What was the primary directive issued by the Oversight Board to Meta?
A: The board mandated the company to remove a video impersonating a woman in an inappropriate sexual context, considering the publication of these fabricated clips to be a non-consensual violation that must be deleted immediately.
Q: What recommendations aim to facilitate the reporting process for abusive content?
A: The board recommended providing dedicated deepfake reporting forms globally and allowing trusted individuals such as family members to submit reports on behalf of targeted victims.
Q: Which social group is most affected by these clips according to the report?
A: The report clarified that the spread of this abusive content disproportionately harms women and girls, as well as ordinary non-celebrity individuals who lack protective tools.
