Meta Urged to Tighten Safeguards Against Intimate Deepfakes Targeting Ukrainians

Meta’s Oversight Board Calls for Stronger Protection Against Sexually Explicit Deepfakes

According to НВ — Техно: Meta’s Oversight Board has recommended that the company expand its adult sexual exploitation policies to include protections for ordinary individuals targeted by sexually explicit deepfakes. This push follows Meta’s failure to remove a deepfake video from Instagram, despite multiple user reports. The board’s recommendations, released on June 24 at 4:30 PM, stress that such content should be treated as non-consensual by default, shifting the burden of proof away from victims. For an English-speaking audience, this case underscores growing global concerns about AI-generated abuse and the accountability of major platforms.

Key Proposals from the Oversight Board

Among the board’s main suggestions are:

• Adding AI-generated fake images and videos to the company’s adult sexual exploitation policy;
• Creating a separate reporting category for sexually explicit deepfakes in complaint and appeal forms;
• Allowing trusted contacts-such as friends or family members-to report violations on behalf of the affected individual;
• Introducing a dedicated complaint form for intimate deepfakes, initially available only to residents of Texas and Florida.

These recommendations emerged from an investigation into a case where Meta failed to act on complaints about a deepfake video. The AI-generated clip showed a woman adjusting her dress, with several frames revealing her underwear. The complaint was filed by a friend of the woman depicted, who had already deactivated her account at the time of the report. Two users flagged the video, but Meta took no removal action.

After the Oversight Board intervened, Meta restricted the video’s visibility for minors but concluded it did not violate its standards enough to warrant removal. The company argued that when the initial complaints were received, it lacked confirmation that the person in the video was a real individual. Had the woman herself reported the clip, it would have been treated as proof of non-consent and grounds for deletion.

The board has required Meta to respond to its recommendations, though the company is not obligated to implement them. In the specific case that triggered the investigation, the board overturned Meta’s decision and demanded the video be taken down. In mid-2025, the board noted that the company’s rule enforcement has been inconsistent. Earlier, in March 2025, the board called for separate policies governing AI-generated content, rather than relying on existing misinformation rules. That call followed the publication of an AI-generated video that falsely showed damaged buildings in Haifa, Israel, posted by an account impersonating a news outlet.

This case highlights rising concerns over the use of artificial intelligence to create content that can damage personal reputations and threaten safety. The Oversight Board’s recommendations may signal a new phase in regulating AI-powered technologies and their societal impact. It also underscores the urgent need for protective mechanisms for victims of sexually explicit deepfakes, which can be weaponized for blackmail or harassment.

As Meta grapples with the implications of deepfake technology, the discovery of a facial recognition feature in its app code for smart glasses raises further questions about privacy and user consent. This development highlights the growing concerns surrounding the intersection of AI and personal data security, making it essential for users to stay informed about how these technologies could impact their lives.