The Hidden Legal Risks of AI-Generated Advertising for Retailers

Generative artificial intelligence is rapidly changing how advertising gets made. What once required days of model shoots, full video crews, and large creative teams can now be produced faster and at lower cost with AI tools. For retailers, that means quicker creation of digital ads, seasonal campaigns, product imagery, social media content, and even updated versions of older brand materials. However, while the efficiency gains are obvious, the legal risks — and responsibility for them — are often minimized.

When entities create AI-generated images, those images may unintentionally reflect copyrighted source material used to train the underlying model. Even when infringement is not obvious on its face, the outputs can still be substantially similar to protected works. For retailers, that creates risk when AI-generated content is used across websites, email marketing, paid media, or national campaigns.

That risk is not theoretical. Copyright owners can use reverse image and audio search tools to identify potentially infringing content after it is published. Retailers should therefore understand how an agency’s or vendor’s AI tools were trained and address the allocation of risk in their contracts. Master services agreements should include clear limitations, indemnities, and content-clearance obligations. In practice, however, the retailer is often the party most likely to face an enforcement claim.

There is also a separate issue: ownership of the AI-assisted content itself. If a retailer wants to protect the value and goodwill invested in a campaign, it must consider whether the resulting work qualifies for copyright protection. U.S. copyright law generally requires meaningful human authorship. Purely machine-generated output usually doesn't qualify.

For that reason, documentation matters. Retailers should keep records of prompts, drafts, edits, sketches, and other evidence of human creative input. Like preserving lab notes in the patent context, maintaining this history can help support claims of authorship and exclusivity. Retailers should also remember that many AI platforms do not treat prompts as confidential, which can undermine efforts to preserve uniqueness.

Trademark law presents a series of added risks. AI systems may generate content that imitates or references protected brand names, logos, slogans, or other source-identifying elements. In retail advertising, where brand identity is central, that can create significant exposure. Before AI-generated materials go live, marketers should clear key names, visual elements, and branding features for possible trademark conflicts. Early review can also help teams refine prompts before problems are built into a campaign.

Rights of publicity and privacy must also be considered. Retailers using AI-generated models, influencers, or lifestyle imagery should be cautious if those images resemble real people or incorporate personal data. It's important to review the AI platform’s disclosures and determine whether protections for publicity and privacy rights exist. If an ad uses or appears to depict an identifiable person, obtaining consent or a release may be necessary. Failing to do so can lead not only to legal claims, but also to reputational damage.

Ultimately, the speed of AI-generated advertising shouldn't replace traditional legal review and brand-protection safeguards. Retailers should implement clear internal procedures governing which tools may be used, what approvals are required, and what contractual and legal checks must occur before publication. A consistent process can help marketing teams move quickly while reducing the risk of copyright, trademark, privacy, and ownership disputes.

Jeremy S. Boczko is a partner in Hunton Andrews Kurth LLP’s intellectual property practice in the firm’s New York office.

Andre Earls is an associate in Hunton’s intellectual property practice in the firm’s Richmond office.