Brands are increasingly using AI-generated models, synthetic performers, voiceovers, avatars and virtual influencers because they can reduce production time and cost. The technology may simplify parts of the creative process, but it does not remove the need for legal clearance. In many cases, it adds a new set of rights, disclosure and contractual issues that should be resolved before the content is released.
New York’s synthetic performer law, which took effect on June 9, 2026, is one example of how quickly the rules are developing. The law requires a conspicuous disclosure when certain advertisements use a “synthetic performer.” It applies when a business knowingly produces or creates an advertisement containing a digitally created performer who appears to be human but is not recognizable as an identifiable real person.
A violation can result in a civil penalty of $1,000 for a first offense and $5,000 for each subsequent violation. The disclosure requirement applies across advertising media, although the law includes exceptions for audio-only advertisements, certain language translations and promotional materials for expressive works when the use is consistent with the underlying work. The New York State legislation does not specify the exact language, placement or format required for a disclosure to be considered “conspicuous,” leaving advertisers to make a reasonable assessment based on the format and presentation of each advertisement.
The law is important, but disclosure is only one part of the review.
Synthetic does not necessarily mean original
A fully generated performer may still incorporate legally significant elements of real people or protected material. An avatar could resemble a working model, celebrity or influencer closely enough to raise right-of-publicity or false endorsement concerns. A synthetic voice may reproduce a recognizable cadence or vocal quality. The content used to direct or build the output may include copyrighted photographs, video, music, scripts or other material that the brand does not have permission to use.
Brands should not assume that an AI vendor’s ability to generate an asset means the resulting content has been cleared. The relevant questions include how the model was created, what reference materials were supplied, whether an identifiable person can be recognized in the result and what rights the vendor grants to the customer.
The analysis also changes when AI is used to modify footage of a real performer. A production agreement permitting ordinary editing may not authorize the creation of a digital replica, an extension of the performance or use of the person’s likeness in entirely new content. Those rights should be addressed expressly rather than inferred from a standard talent release.
Vendor agreements require closer review
AI platform terms vary significantly. Some vendors provide broad commercial rights in generated outputs, while others reserve rights to reuse prompts, uploaded assets or generated content. Contract terms may also contain limited warranties, broad disclaimers and liability provisions that leave the brand responsible if a third party challenges the advertisement.
Before using AI-generated talent, brands should confirm:
- Who owns or may commercially use the final output
- Whether prompts, reference images and uploaded materials may be retained or used for training
- Whether the vendor provides warranties concerning its technology and training materials
- Who is responsible for claims involving copyright, trademark, publicity rights or consumer deception
- Whether the content can be modified, licensed and used in every intended market
Agencies should address these issues with clients before production begins. A brand may approve the creative concept without realizing that the proposed workflow uses synthetic talent or that the agency’s vendor terms do not provide the rights expected by the client.
Clearance should begin before generation
The most efficient approach is to treat AI talent as part of the production clearance process, not as a shortcut around it. Before generating content, the brand and its agency should determine whether a synthetic performer is appropriate, identify the tools and source materials that will be used, and establish any restrictions on appearance, voice or conduct.
The final content should then be reviewed for recognizable people, third-party intellectual property, required disclosures and advertising claims. Brands should also retain records showing how the asset was created, which materials were provided to the system, what approvals were obtained and which version was ultimately published.
New York’s law does not prohibit synthetic performers. It requires transparency when covered content is used in advertising. For brands, the broader lesson is that AI may change how talent is created, but it does not eliminate the rights and responsibilities involved in commercial production. Legal review should evolve with the technology rather than disappear from the process.
