Deepfake marketing sits at the intersection of synthetic media, advertising, and trust. AI can now alter a person’s face or voice convincingly enough to create new video, audio, and interactive experiences. For brands, that opens creative possibilities, but it also creates a simple risk: if audiences believe a real person said or endorsed something they never approved, the campaign has crossed from synthetic media into deception.
The better way to think about deepfake marketing is as a consent and disclosure problem before it is a production technique. A legitimate marketing campaign should have clear rights to the face, voice, or identity being synthesized; truthful claims; and enough disclosure that a reasonable viewer is not tricked about who actually appeared or spoke.
What Is Deepfake Marketing?
A deepfake is synthetic or heavily manipulated media created with machine learning or generative AI to make a person appear to say or do something that did not occur in the original recording. The same underlying techniques can be used for face replacement, voice cloning, virtual presenters, translated performances, digital doubles, and interactive characters.
Not every synthetic video is deceptive. A disclosed digital avatar made with permission is very different from a fake celebrity endorsement. That distinction matters for entertainment marketing, advertising, museums, ecommerce, and branded content, where synthetic media may be part of the creative format rather than an attempt to impersonate someone.
A Positive Use Case: Dalí Lives
The Dalí Museum offers one of the clearer examples of synthetic likeness used with an educational purpose. In 2019, the museum launched Dalí Lives, an AI experience that recreated Salvador Dalí’s likeness on interactive screens. The museum says the project used thousands of frames, archival interviews and writings, an actor, and machine learning to create new footage designed to help visitors engage with Dalí’s life and work.
The marketing lesson is that the technology served the visitor experience. The synthetic likeness had a defined context, the institution explained what it was, and the creative concept matched Dalí’s legacy. This is where user experience matters: the technology should make the interaction clearer or more engaging rather than existing only to create novelty.
Fake Celebrity Endorsements Are Not a Marketing Shortcut
The more damaging use of deepfake advertising is the unauthorized endorsement. Scammers have used altered video and cloned voices to make celebrities appear to promote cookware giveaways, supplements, investments, and other offers they never endorsed. The U.S. Federal Trade Commission has specifically warned consumers about fake celebrity and influencer endorsements made with doctored video and audio.
For legitimate advertisers, the baseline is straightforward. The FTC’s endorsement guidance says endorsements must be truthful and not misleading, and a celebrity endorsement must reflect the celebrity’s honest experience or opinion. A synthetic endorsement does not create an exception to those rules. If the person did not authorize or genuinely make the endorsement, the ad should not present it as if they did.
Taylor Swift, Selena Gomez, Elon Musk, and Joe Rogan
High-profile people including Taylor Swift, Selena Gomez, Elon Musk, and Joe Rogan have all been used in deceptive synthetic-media schemes or fake endorsement content. These incidents are better described as impersonation or fraud than as successful deepfake marketing. The point for brands is not which celebrity was targeted; it is that recognizable likenesses can transfer trust to a scam quickly.
That risk extends to the companies associated with the impersonated person. Fraudulent content can create customer-service problems, reputation damage, false claims, and confusion about whether a brand, founder, or spokesperson actually approved the message. Protecting brand trust therefore includes monitoring for impersonation as well as controlling what the company publishes itself. An AI image detector can be one part of that review, alongside source verification and human judgment; its output should not be treated as proof that an image is authentic or manipulated.
The Benefits and Risks of Synthetic Media in Marketing
Where It Can Help
Zalando’s current experimentation with digital twins of real models is a useful modern example. Zalando describes the models as high-fidelity replicas developed with the people involved to make content production more flexible while maintaining accuracy and human oversight. That is closer to responsible synthetic production than deceptive impersonation.
Where It Can Go Wrong
Disclosure Rules Are Becoming Part of the Creative Brief
Major video platforms now treat realistic synthetic media as something audiences should be able to identify. YouTube requires creators to disclose altered or synthetic content that looks realistic. TikTok’s advertising policy requires significantly AI-modified media to be labeled and says ads may not misuse the likeness or voice of a public figure without permission, including false endorsements.
For a brand, that means disclosure cannot be added at the last minute. The production brief should document whose likeness or voice is being used, what consent exists, whether the content changes the meaning of an original performance, how it will be labeled, and what each platform requires.
The creative also needs to remain strong after disclosure. If a campaign only works when the audience thinks the synthetic scene is real, that is a warning sign rather than a reason to hide the label.
Impersonation Scams Are a Growing Brand-Safety Issue
FTC data show that reported losses to imposter scams reached $3.5 billion in 2025. That figure covers many kinds of impersonation, not only deepfakes, but it shows why synthetic identity fraud should be treated as a serious risk rather than an unusual edge case.
The FTC’s federal Impersonation Rule currently gives the agency stronger tools against government and business impersonation. The agency has also proposed extending the rule to impersonation of individuals, but that proposed individual-impersonation amendment should not be described as a finalized rule. State publicity, privacy, fraud, contract, and other laws can also apply depending on the use and jurisdiction.
This article is not legal advice. For any campaign that copies a real person’s face, voice, performance, or identity, brands should obtain appropriate legal review before publication.
A Practical Deepfake Marketing Checklist
The Future of Deepfake Marketing Is Really About Trust
Synthetic media will become easier to produce, but that does not mean audiences will accept every use. The competitive advantage is likely to come from brands that can use AI without making customers wonder whether the company is hiding who created, approved, or performed the message.
That makes campaign design, disclosure, rights management, and governance part of the same creative process. The technology can lower production friction, but it cannot replace judgment about whether a campaign is fair, clear, and consistent with the brand.
Brand Vision can help teams evaluate where synthetic media fits within a broader campaign, how it affects brand trust, and which customer touchpoints need human review. A marketing consultation can help define those guardrails before production begins.
