A Deepfake Taylor Swift Ran a Cookware Giveaway Scam on Her Own Fans
What happened
In January 2024, AI-generated video ads began circulating on Facebook and TikTok featuring what appeared to be Taylor Swift announcing a free Le Creuset cookware giveaway. Swift had no involvement. Le Creuset had no involvement. The videos were fabricated using AI tools that synthesized her likeness and voice, then deployed as paid advertisements to funnel viewers into a scam designed to steal money and personal data from the fans who trusted what they saw.
The script followed a social media giveaway template close enough to read as plausible. The synthetic Swift told viewers that a packaging error had left 3,000 cookware sets unsellable and she was giving them away to loyal fans for free. Users who clicked through were taken to survey pages, the standard funnel for extracting payment details or personal information under the pretense of claiming a prize. The content was polished enough that the premise, a celebrity giveaway tied to a logistics quirk, did not immediately signal fraud on a fast-moving feed.
The campaign ran across at least Facebook and TikTok, reaching audiences well beyond any organic channel. Both Swift and Le Creuset were used as props without consent: Swift for her reach and the devotion of her fanbase, Le Creuset for the credibility attached to its brand. The scammers needed neither party to cooperate. They only needed AI tools capable of producing a convincing enough likeness and platforms willing to run paid ads before verifying the identity or authorization of whoever was buying the placement.
This kind of operation does not require sophisticated resources. Commercial tools capable of voice synthesis and video manipulation are widely available, and deploying them against a recognizable celebrity with a large and loyal following is cheap. Swift was an obvious target, one of the most recognized faces in the world at that moment, with a fanbase whose enthusiasm scammers could count on to outrun skepticism. The same template has been applied to other public figures using the same basic mechanics, and nothing about the platforms' advertising systems made it structurally harder to run the second time than the first.
The scammers behind the ads were never identified. No arrest, no platform enforcement action, and no public attribution followed the reporting. That absence points to a gap that runs deeper than any single moderation policy: there is no standard mechanism for proving who authorized a piece of synthetic content before it is accepted for paid distribution. A provable record of what a system did, who submitted the output, and what authorization it carried would have made the fraud detectable at the point of upload rather than after money and data had already moved. Without that record, any actor can borrow a public figure's face, run it through an ad network, and be gone before the damage is counted.
Reported impact
- Affected parties
- Not publicly disclosed
- Harm type
- Not publicly disclosed
- Scale
- Not publicly disclosed
- Financial impact
- Not publicly disclosed
- Regulatory action
- Not publicly disclosed
Classification
Relevant governance controls
Governance control mapping is not available for this record.
- No controls mapped
Not publicly disclosed
Control mapping is analytical. It does not state that any control would have prevented the incident.
Sources and evidence
This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.