Unfair Competition FTC Enforcement Against AI

Some of the world’s largest companies have rolled out their newest large language models, allowing developers and the public to use the technology firsthand. AI has changed the way businesses operate. Tasks that once took hours now take minutes.

All new technologies raise questions about the proper role of government oversight. Two regulatory approaches have emerged. The EU favors the ex ante approach, which entails regulating AI before harm occurs.

Alternatively, a jurisdiction might use the ex post approach because premature regulation risks stifling innovation and imposing high costs without a clear public benefit.

Enforcement moves by the FTC offer an early look into the direction the country could take. In recent cases, the FTC stretched its authority under the FTC Act to target not just deceptive claims, but also the potential misuse of neutral AI tools by others.

The FTC Act makes any unfair or deceptive acts or practices unlawful.

Congress left it broad, recognizing that a consumer protection law can't cover all unfair trade practices. The FTC can't use the APA for rulemaking, unlike other agencies. Instead, it relies on the more complex Magnuson-Moss Act rules, which require extra steps. Usually, the FTC relies on guidance, court decisions, consent orders, and the substantial-injury test to define its limits. Over time, some clarity has emerged on how the FTC uses its authority: (1) deceptive practices and (2) unfair practices.

To prove a deceptive act or practice, the FTC must show there was [1] a representation, omission, or practice, that [2] is likely to mislead consumers acting reasonably under the circumstances, and [3] the representation, omission, or practice is material.

The FTC periodically issues rules or guidance on these factors, sometimes on its own initiative or under Congress's authority for expedited rulemaking. For example, FTC rules clarify when a "Made in the USA" claim is deceptive, and the Green Guides help companies assess if their environmental claims are deceptive. Congress directed the agency to clarify "deception” in “Made in USA" claims and granted rulemaking authority, but the agency also provided clear guidance on environmental marketing claims independently and under existing authority.

Like the deception prong, the unfairness prong is a broad prohibition; however, unfairness claims benefit from additional statutory clarity provided by Congress.

An act is unfair if it causes or is likely to cause substantial injury to consumers that is not reasonably avoidable and not outweighed by benefits to consumers or competition. The FTC says that substantial injury means more than trivial harm, often requiring monetary injury or health or safety risks. The injury must be unavoidable, and the marketplace should be self-correcting through consumer choice. When sales techniques prevent consumers from making effective decisions, the FTC is more likely to intervene.

The unfairness test also requires examining whether the disputed practice violates public policy.

The Commission considers First Amendment decisions to determine if restricting advertising hampers informed consumer choice. Public policy can support an FTC unfairness action, such as when a mail-order company filed collection suits in an inconvenient forum. When the FTC relies on public policy, it is clear and well-established.

Congress empowered the FTC with broad enforcement authority. Congress intentionally granted broad consumer-protection powers to prevent the FTC’s authority from being blunted, while safeguarding the flexibility of commercial activity to serve consumers’ needs through newer technologies and innovative methods.

The FTC employs its enforcement powers against both outdated and modern unfair and deceptive practices, from false advertising in newspapers to deceptive AI platforms, while avoiding cases in which newer technologies may pose only theoretical risks of harm.

Beyond targeting direct actors, the FTC can also hold companies responsible for enabling others’ deception by providing the tools that enable it.

This liability is based on the premise that the originator who provides the means and instruments to a third party that then deceives consumers or engages in unfair practices remains culpable under the FTC Act. The FTC first introduced this form of liability in a case in which a manufacturer that did not sell directly to consumers violated the law by falsely advertising that certain garments were made entirely of wool when they were not. The case did not involve a claim of unfair or deceptive sales practices but rather a claim of unfair competition, as the false advertising allegedly diminished the market for the sale of honest goods. The U.S. Supreme Court concluded that, even though the manufacturer was not directly selling to the public, it was still liable because it furnished retailers with the means to commit fraud on the public.

This doctrine is not limited to claims of unfair competition. In another case, the court found that the defendant operated a pyramid scheme and made false and deceptive claims that violated the FTC Act. The court agreed with the FTC that the defendant had provided the means to others to perpetuate the pyramid scheme, finding those at the top of the pyramid liable for deceiving the other members of the operation.

The FTC has also resorted to this doctrine when companies use misleading advertising. A manufacturer that provided retailers with fictious suggested list prices but that represented that those prices were customary retail prices was liable for unfair or deceptive acts, even though it never dealt directly with the public. An appeals court affirmed that a clockmaker was liable for licensing its trademark to a company it knew was deceiving consumers with the trademark.

Traditionally, the means doctrine was reserved for two types of cases: where a defendant supplied a retailer with a service or product that violated the law, and where a defendant sent misleading or fraudulent marketing materials down a supply chain to consumers.

To establish means liability, the FTC must satisfy the three-part test: [1] a representation, omission, or practice, that [2] is likely to mislead consumers acting reasonably under the circumstances, and [3] the representation, omission, or practice is material, and then it must prove that the originator had knowledge or reason to expect that consumers may possibly be deceived as a result of its false or misleading actions.

The FTC first successfully applied the means doctrine to an AI tool in 2024. It brought an enforcement action for a violation through an AI-enabled writing assistant service. The company provided its clients with 43 distinct use cases, including Email, Product Description, and blogs, to help generate written content for their websites.

It was a Testimonial & Review service that drew FTC scrutiny. The service allowed clients to use AI-generated written content for product reviews. Clients enter product keywords, choose a tone and language, and receive tailored review content. The generated content had no relation to the input, leading to reviews that would almost certainly be false for users who copied it and published it online. Despite the legitimate uses, the FTC concluded that because the service can quickly generate thousands of reviews with minimal input, its likely only use is to facilitate subscribers' posting of fake reviews to deceive consumers. 24 users generated over 10,000 reviews each, and 114 users generated over 1,000 reviews each. The FTC alleged the service lacked legitimate use and that the harm outweighed any public benefit. The FTC claimed that because the company furnished its clients with the means to generate false and deceptive reviews, it was liable for providing the means to commit deceptive acts and practices.

The FTC successfully imposed means liability on a company for an AI tool it advertised as an AI-based review platform.

The website featured more than 130,000 businesses, each with a star-rating review system linked to customer reviews. However, most of the customer reviews were from instant surveys. It offered two types of instant surveys: Instant Feedback Surveys (IFS) and Instant Feedback Product Reviews (IFPR) through its business-facing platform, Jabio. Jabio marketed itself as an artificial intelligence-enabled suite of tools for businesses to gather and manage customer reviews. The IFS is an automated survey that pops up after the consumer completes a transaction on a client’s website, asking consumers to rate their overall shopping experience so far on a scale of zero to five stars. When pitching IFS to clients, the company explained that the survey was designed to capture customers at their happiest and most engaged. The IFS also generated an unprecedented volume of positive feedback to keep a client’s profile at an impeccable score.

Similarly, the IFPR survey automatically popped up at the point of sale and asked the consumer, “Why did you choose this product today?”

The review prompted the consumer to rate the product on a scale of 0 to 5 stars. The data was used to generate publicly accessible product ratings on the client’s profile page. It also provided clients with product review widgets that enabled them to display IFPRs on their own websites. Third parties used the widget to display IFPRs in Google’s paid product search results, allowing clients to show their product reviews before a consumer even clicked on their website.

For both the IFS and IFPR services, clients’ customers were invited to leave reviews before receiving the product or service. It was falsely represented that the reviews were from consumers who had received the product or service and had the opportunity to experience it themselves.

The FTC claimed that these instant surveys artificially inflated the volume of reviews and the average review ratings for clients. The FTC also brought a claim that the company provided the means to engage in deceptive practices by offering clients a widget to display the IFPR rating on their websites.

Since June 2024, the FTC has filed multiple complaints alleging violations in AI-related cases.

Recent decisions, commentary, and public statements offer clues about the FTC’s expected enforcement posture: cautious optimism about AI’s promise, paired with a focus on encouraging innovation. Even if the Commission’s membership shifts in 2026, the analytical approaches reflected in its published opinions will likely continue to guide the agency's application of the FTC Act to AI. The Commission is also likely to limit unfairness claims to fact patterns in which it can clearly show that it has met the statutory substantial injury balancing test. 

The takeaway is that the FTC is unlikely to bring M&I claims solely for offering neutral tools. Liability is more likely to fall on companies that use AI to commit fraud or deception, rather than on those that develop platforms with lawful, general-purpose applications that are misused by third parties.