Deepfake Impersonation Advertising

Hello, this is Legal Agent.

Generative AI has made it easy to imitate a person's face, voice and manner of speaking, and impersonation of celebrities, executives, doctors and lawyers has started showing up in ads and social posts. Startups should treat this as more than a risk of being caught up in someone else's bad ad. It also raises the question of how far a company can go when using another person's likeness, name or title in its own advertising.

Risk in using someone's face, voice or name

Using another person's face, voice or title in an advertisement generally requires consent. Content that makes it look as though an impersonated person is endorsing a product raises misrepresentation risk under the Act against Unjustifiable Premiums and Misleading Representations, and using a person's likeness or voice without consent can itself infringe rights of publicity and portrait rights, alongside possible defamation, copyright or trademark issues. Impersonating a company name or logo can also draw a claim under the Unfair Competition Prevention Act as misappropriation of a well-known indication. Even an AI-generated person can raise concern if it resembles a real individual. An ad resembling a celebrity, doctor or investor risks consumers assuming that person's endorsement, so borrowing a third party's credibility calls for checking consent, disclosure and contract scope.

First response if your company is impersonated

If a company or its executives are impersonated, preserving evidence comes first: URL, screenshots, ad ID, timestamp and the destination link, before pursuing platform takedown requests, identifying the advertiser, contacting police or issuing a public warning. Deleting in haste, before evidence is preserved, can make later tracing difficult.

Managing the risk through agency contracts

Contracts with advertising agencies and production companies should cover rights clearance for ad materials, whether AI-generated assets are permitted, and liability if something goes wrong, including which tools an agency uses and whether the output carries third-party rights risk. Advertising built around a person of authority in finance, medical or investment contexts warrants particular care given its influence on consumers.

Internal communications need AI rules too

Internal and recruitment communications also use AI-generated images and voice: an employee-like portrait, a synthesized voice for an executive, an event photo that never happened. Presenting these as if they were real can mislead, so companies should decide whether to label AI-generated material or phrase it so it cannot be mistaken for a real person. Beyond copyright, PR-facing legal review should also ask whether the material could undermine trust in what the company communicates.