AI Brand Infringement
& Deepfake Defense

Patterns include deepfake videos showing you saying or doing things you never did, AI voice clones used in advertisements or product endorsements you never approved, fake brand impersonation accounts on social platforms, and AI-generated logos or marketing content that copies your registered trademarks. The legal frame depends on which element was misused: federal trademark law protects registered marks, and 15 U.S.C. § 1125(a) also reaches, with or without a registration, uses likely to cause confusion as to your affiliation with, or sponsorship or approval of, another's goods, services, or commercial activities; California's right of publicity statute addresses specified unconsented commercial uses of your name, voice, signature, photograph, or likeness, subject to its exceptions; and state-specific deepfake statutes layer on top.

In California, subdivision (a)(1) of Civil Code Section 3344 imposes liability for injury caused by specified knowing uses of another person's name, voice, signature, photograph, or likeness on or in products, merchandise, or goods, or for advertising, selling, or soliciting purchases, without prior consent. The liable person owes the greater of $750 or actual damages, plus profits attributable to the use that are not already taken into account in actual damages; punitive damages may also be awarded, and the prevailing party is entitled to attorney's fees and costs. Subdivision (d) excludes uses connected with news, public affairs, sports broadcasts or accounts, and political campaigns; subdivisions (e) and (f) further limit liability for uses in commercial media and for owners or employees of advertising media. The remedy depends on which state's law applies.

It depends on the deepfake. California Penal Code § 647(j)(4)(A)(ii) makes it a misdemeanor to intentionally create and distribute, or cause to be distributed, a realistic computer-generated intimate image of an identifiable person, created so that a reasonable person would believe it authentic, when the distributor knows or should know that distribution will cause serious emotional distress and the person depicted suffers that distress. That clause does not apply to a person who was under 18 when committing the offense. Subparagraph (D) excludes distribution in reporting unlawful activity, in compliance with a subpoena or court order for a legal proceeding, during a lawful public proceeding, or related to a matter of public concern or public interest; public-figure status alone does not establish that last exception. Such images can also be civilly actionable under Civil Code Section 1708.86, under which a prevailing depicted individual who suffers harm may elect, at any time before final judgment, statutory damages of $1,500 to $50,000 for all unauthorized acts involved in the action with respect to any one work, or up to $250,000 if the act was committed with malice. At the federal level, the TAKE IT DOWN Act makes it unlawful, under conditions the statute lists, to knowingly publish intimate visual depictions or AI-generated digital forgeries of an identifiable individual without consent, and requires covered platforms to remove such a depiction within 48 hours of a valid removal request from the individual or an authorized person. In August 2025, in Kohls v. Bonta, a federal district court permanently enjoined the California Attorney General and Secretary of State from enforcing two California election-deepfake statutes: AB 2839 against the named plaintiffs, and AB 2655 against X Corp. and Rumble. Its August 28 order, filed August 29, barred AB 2655 enforcement against any provider of an interactive computer service and provided that it would terminate if an appellate court vacated the AB 2655 judgment. Non-sexual, non-election deepfakes are not categorically illegal, but they may be actionable under existing torts including defamation, right of publicity, and fraud.

The NO FAKES Act is bipartisan federal legislation that has not been enacted; as of November 3, 2026, the 2026 version (S. 4591) had been reported by the Senate Judiciary Committee (June 24, 2026) and was on the Senate Legislative Calendar. The NO FAKES Act link points to the 2025 version (S. 1367); the bill reported in 2026 is S. 4591. Until a bill is enacted, a federal registration of your name, signature image, or sound mark gives you a civil action under 15 U.S.C. § 1114(1) against confusingly similar use, and 15 U.S.C. § 1125(a) claims do not require a registration.

Yes for marks that function as source identifiers in commerce; not directly for your face or unrecorded voice. TMEP § 1202.15 permits registration of sounds that identify and distinguish the source of goods or services. On December 16, 2025, the USPTO registered to J.K. Livin Brands, Inc. a sound mark consisting of a man saying ALRIGHT ALRIGHT ALRIGHT (Reg. No. 8,070,191). You cannot trademark your face in the abstract, but you can register a stylized image that functions as a logo. Lanham Act § 43(a) provides a federal civil action, with or without a registration, against uses likely to cause confusion as to your affiliation with, or sponsorship or approval of, another's goods, services, or commercial activities.

We recommend the following response steps, with sequencing tailored to the incident. First, preserve evidence: full screenshots of the content, the URL, the platform timestamp, and any metadata available before the post is taken down. Second, identify every platform hosting the content. Third, we recommend retaining counsel before sending any direct communication to the creator. Fourth, file attorney-drafted DMCA takedowns where copyright applies, platform-specific impersonation reports to Meta, X, TikTok, and YouTube, and a right-of-publicity demand to the underlying creator account. Sequencing those steps is what the strategy call exists to set.

Trademark protects source identifiers used in commerce, including your name when it functions as a brand, distinctive logos, signature images registered as marks, and sound marks. Copyright protects original creative works fixed in a tangible medium, such as your photographs, videos, written content, and audio recordings, but copyright does not protect your name, likeness, or persona in the abstract. Right of publicity, codified for example at California Civil Code Section 3344, imposes liability, subject to that section's conditions and exceptions, for specified knowing uses of your name, voice, signature, photograph, or likeness on goods or for advertising or selling without your prior consent, whether or not you have registered anything. A complete protection stack uses all three: trademarks for registered-mark claims under the Lanham Act, copyrights for content you create, and right of publicity for unregistered identity elements.

Under 47 U.S.C. § 230(c)(1), no provider or user of an interactive computer service shall be treated as the publisher or speaker of information provided by another information content provider. Section 230(e)(1) preserves the enforcement of federal criminal statutes, and section 230(e)(2) provides that nothing in the section shall be construed to limit or expand any law pertaining to intellectual property. Federal courts of appeals read section 230(e)(2) differently: the Ninth Circuit construed intellectual property to mean federal intellectual property (Perfect 10, Inc. v. CCBill LLC, 2007), while the Third Circuit held that the provision is not limited to federal law and that a Pennsylvania statutory right-of-publicity claim against Facebook arose out of a law pertaining to intellectual property (Hepp v. Facebook, 2021). Under California Business and Professions Code § 22671 (SB 981), a social media platform must give California-resident account holders a way to report sexually explicit digital identity theft, temporarily block a reported instance pending its determination, and remove it immediately if it determines there is a reasonable basis to believe the report is valid. The TAKE IT DOWN Act requires covered platforms to remove a non-consensual intimate visual depiction within 48 hours after a valid removal request.

Enforcement is split across multiple authorities depending on the type of misuse. A public prosecutor, a term that includes the California Attorney General, may bring a civil action to enforce Civil Code § 1708.86. AB 2602 (Labor Code § 927) makes a provision in an agreement for personal or professional services unenforceable as to a new performance by a digital replica fixed on or after January 1, 2025, when the statute's conditions are met. The Federal Trade Commission enforces against deceptive AI-generated impersonation under Section 5 of the FTC Act. Civil Code § 1708.86 (as amended by AB 621) gives a depicted individual a cause of action, and Civil Code § 3344.1(a)(2) provides a remedy for specified unauthorized digital replicas of deceased personalities in expressive audiovisual works or sound recordings, subject to its defined terms, consent rules, and express exceptions for news, public affairs, sports, comment, criticism, scholarship, satire, parody, qualifying documentary or historical or biographical representations, fleeting or incidental uses, and advertisements for the excepted works. Under 15 U.S.C. § 1121(a), federal district courts have original jurisdiction of all actions arising under the Lanham Act, including § 1125(a) claims, which do not require a registration.

We recommend the following preparation steps. Register the federal trademarks that match your AI-exposure surface, including your name as a brand mark, your signature image as a logo, and a sound mark for any catchphrase that identifies you commercially. Update vendor, employment, and licensing contracts to address AI-generated content directly, including no-AI-replication clauses and AI-training-data restrictions. Set up monitoring infrastructure across the platforms most likely to host AI infringement of your brand. Document the legitimate baseline of your voice, image, and video assets so you have a reference point if synthetic media surfaces. We recommend building the incident response plan in advance.