Media

Actors say AI sex ads hijack their faces

Micro drama marketers push fake scenes on TikTok and Meta, ad approval scales faster than identity disputes

Images

From left: vertical actors Tess Dinerstein, David Eves, and Faith Orta
                              
                                Gregory Wallace, Yellowbelly, Yellowbelly From left: vertical actors Tess Dinerstein, David Eves, and Faith Orta Gregory Wallace, Yellowbelly, Yellowbelly businessinsider.com

A set of AI-generated ads circulating on TikTok and Meta has been using actors’ faces in fake sex scenes to sell “micro drama” apps, according to Business Insider. Actors including Tess Dinerstein, David Eves and Faith Orta told the outlet they discovered their likenesses in explicit clips that they say they never filmed and that do not appear in the shows being advertised. The ads, they said, were presented as if they were excerpts from real series rather than synthetic bait.

The micro-drama business runs on cheap production and expensive distribution: short episodes designed for mobile viewing, pushed through performance marketing on the same platforms that also host the audience. That model rewards whatever stops a thumb-scroll, and it punishes whoever tries to verify provenance, because verification slows down ad buying. A fake clip that converts can be uploaded, tested, and replaced faster than a takedown request can be processed, and the cost of being wrong is mostly borne by the person whose face is used.

For platforms, the incentives are similarly lopsided. Ad systems are built to approve at scale and to catch obvious fraud, not to authenticate every human depicted in every frame. The more automated the pipeline, the more valuable it becomes to adversaries who can generate endless variants—slightly different edits, crops, or overlays—to evade matching tools and moderation queues. Even when a platform removes one ad, the campaign logic remains: the buyer learns what passed, tweaks the prompt, and tries again.

For performers, the harm is not just embarrassment but attribution. Casting decisions and reputations can turn on a single viral clip, and the actor has no way to prove a negative to viewers who see their face attached to a scene. Legal remedies exist in theory—right of publicity, defamation, platform reporting channels—but they operate on timelines measured in weeks and months, while ad campaigns run in hours. In practice, the fastest actor response is often to do unpaid investigative work: screenshotting, filing reports, and contacting journalists.

The episode also underlines how synthetic media is being normalised first in the least regulated corner of the market: low-budget apps that buy reach rather than build brands. A studio that expects to work with the same talent and distributors next year has reasons to avoid reputational landmines. An app that can churn through titles and ad accounts can treat complaints as a cost of acquisition.

Business Insider’s reporting names specific actors and describes specific ads, but the companies behind the campaigns can change faster than the clips themselves. The platforms still sold the placements while the performers did the cleanup.