Key points

• Existing advertising law applies to AI-generated content. Claims must be truthful and substantiated whether written by a person or a model, and the advertiser is responsible.
• The FTC's rule on consumer reviews and testimonials prohibits fake reviews, including those generated by AI, and the FTC has brought cases against companies making unsupported claims about AI itself.
• Synthetic people, voices and testimonials raise misrepresentation, right of publicity and, in some states, specific disclosure obligations.
• The EU AI Act's transparency obligations for synthetic content are scheduled to apply from August 2026, although the Commission has proposed adjustments, so timing and technical detail should be checked.
• A short internal policy on approvals, human review, records and disclosure will prevent most problems and is quick to adopt.

Why this is now a legal department issue

Marketing teams have adopted generative tools for copy, product images, video and personalized messages. The speed gain is real, and so is the risk that content goes live without the review that traditional campaigns receive. When an AI-generated image shows a product performing better than it does, or a generated testimonial describes an experience that never happened, the legal analysis is not novel. The risk lies in the volume and the lack of a checkpoint.

Disclosure and deception

In the United States, Section 5 of the FTC Act prohibits deceptive advertising and applies regardless of the technology used. The agency has said that the use of AI does not create an exemption, and its enforcement program has targeted businesses that make exaggerated or unsupported claims about what their AI products can do. For advertisers, the practical questions are whether a reasonable consumer would be misled by synthetic content, and whether a disclosure is needed to avoid that. Realistic depictions of people who do not exist, or of real people saying things they did not say, are the clearest cases.

The FTC's Consumer Reviews and Testimonials Rule, effective since late 2024, prohibits fake or false reviews and testimonials, including those that misrepresent that they come from a person who does not exist or who had no actual experience, and it expressly reaches AI-generated content. Civil penalties are available. Brands should ensure that any use of generated reviews, influencer content or customer stories is prohibited unless clearly identified, and that agencies and affiliates are bound by the same standard.

The Lanham Act and state laws also matter. Competitors can sue over false advertising, and the National Advertising Division, which operates a self-regulatory process, has reviewed claims in AI-assisted campaigns.

Substantiation

Advertisers must have a reasonable basis for objective claims before they are made. Generative tools can produce confident statements about efficacy, ingredients, sustainability or performance that have no source. Review processes should require that every factual or comparative claim be matched to evidence in the file, regardless of how the copy was created. This is particularly important in regulated product categories such as food, cosmetics, supplements and children's products.

Likeness, voice and intellectual property

Using a synthetic version of a real person's face or voice without consent can lead to claims under state right of publicity laws, and several states have enacted statutes specifically addressing digital replicas, including Tennessee and California. Contracts with talent should address the use of AI to modify or recreate their likeness. Copyright ownership of generated material is also unsettled in the United States, where purely machine-generated output generally cannot be registered, so brands should consider what they own in campaign assets.

United Kingdom and European Union

The UK Advertising Standards Authority has said that its codes apply to AI-generated advertising and has asked advertisers to avoid misleading uses of the technology. The CMA's enforcement powers under the Digital Markets, Competition and Consumers Act, including the ban on fake reviews, apply equally. In the EU, the Unfair Commercial Practices Directive covers misleading content, and Article 50 of the AI Act requires labeling of certain synthetic audio, image, video and text content, and disclosure of deepfakes. The Commission has proposed changes to the timing of some obligations, and codes of practice are expected on marking and labeling, so brands operating in Europe should monitor developments.

A practical policy

• Identify approved tools and the data permitted as input, including restrictions on confidential product and customer data.
• Require human review and legal sign-off for claims, comparisons and any depiction of people.
• Prohibit synthetic reviews, testimonials and endorsements unless disclosed, and bind agencies to the same rule.
• Keep a record of prompts, outputs and approvals for high-risk campaigns, so that substantiation can be shown.
• Add AI-specific terms to agency, influencer and talent agreements covering disclosure, likeness and indemnities.
• Plan labeling for synthetic media in markets where it is or will be required.

Brands that set expectations early will be able to scale generative tools with confidence. Those that do not often discover problems only after a competitor complaint, a regulator letter or social media reaction.