False Promises in the Age of AI: How the FTC Polices Deceptive Marketing Claims
Victoria Jung
How much can you trust a company that advertises AI as “unbiased,” “fully automated,” or “guaranteed to improve outcomes”? Since ChatGPT burst onto the scene in late 2022, generative artificial intelligence has taken the marketplace by storm.[1] As businesses scramble to embrace the technology, many have pivoted to marketing AI-powered products that promise to revolutionize everything from legal services to e-commerce.[2]
However, as many of us in the legal and tech sectors have observed, this “AI hype” often outpaces reality. The Federal Trade Commission (FTC) has been watching closely, making it clear that new technology does not create a new legal standard for truth in advertising.[3] In fact, FTC Chair Lina M. Khan recently emphasized that there is “no AI exemption from the laws on the books.”[4] While some frame AI-assisted marketing as a novel policy dilemma, the FTC already possesses robust authority under Section 5 of the FTC Act to regulate these claims, and its recent “Operation AI Comply” enforcement sweep proves that traditional deception standards apply regardless of how “emerging” the technology may feel.[5]
At its core, Section 5(a) of the FTC Act prohibits unfair or deceptive acts or practices in or affecting commerce.[6] To the FTC, a representation, omission, or practice is deceptive if it meets a specific three-part test: (1) It is likely to mislead a reasonable consumer; (2) The consumer’s interpretation is reasonable under the circumstances; and (3) The misleading claim is material, meaning it is likely to affect a consumer’s choice regarding a product.[7] Crucially, from a liability standpoint, no intent to deceive is required.[8] The standard applies to both explicit and implicit claims, as well as marketing written by a human or generated by the very AI tool being sold.[9]
One of the most important takeaways for businesses today is that the FTC is not treating AI claims as “one-offs.” Instead, there is a pattern of enforcement focused on a central legal requirement: Substantiation.[10] Advertisers must possess “competent and reliable evidence” to support any objective claims about a product’s accuracy, efficiency, or decision-making capabilities.[11]
Recent cases highlight the high stakes of failing this requirement: In January 2025, software provider, accessible, was required to pay $1 million to settle allegations that it misrepresented its AI-powered “accessWidget.”[12] The company claimed the tool could make any website fully compliant with Web Content Accessibility Guidelines (WCAG), a claim the FTC found to be false or unsubstantiated.[13] In March 2025, Cleo AI agreed to pay $17 million to resolve allegations that it misled consumers about cash advances.[14] The FTC alleged the company sued an AI risk classifier to determine advances but failed to disclose material information about the timing and amounts of those funds.[15]
Deception in this space often falls into a few predictable traps. For example, claiming “our AI eliminates bias” is an incredibly high bar. The FTC took action against IntelliVision Technologies for falsely claiming its facial recognition software was free of racial or gender bias when, in reality, it lacked the evidence to support such a claim and was trained on a much smaller dataset than advertised.[16] Similarly, the FTC previously highlighted concerns with Rite Aid, alleging the retailer used AI facial recognition for surveillance without reasonable safeguards, leading to biased results.[17] Other common deceptions include: overstating accuracy[18] and “fully automated” dreams,[19] among many others.
For business leaders and in‑house counsel navigating this space, the FTC’s message is unmistakable: Innovation does not excuse overstatement. To safeguard against liability, experts recommend:
1. Audit AI Marketing Claims: Review all scripts and advertisements to ensure they accurately represent current capabilities.[20]
2. Substantiate Before Publishing: Data must be substantiated before claims such as a 99% accuracy of “bias-free results” are announced.[21]
3. Avoid Absolute Language: Terms like “guaranteed,” “perfect,” or “fully autonomous” are red flags for regulators.[22]
4. Disclose the “Human in the Loop”: If the “automated” system requires significant human intervention, failing to disclose that is a material omission.[23]
5. Be Transparent About Training Data: Do not claim the model is trained on “millions” of data points if the reality is much smaller.[24]
While the technology under the hood of our products is evolving at a breakneck pace, the legal framework governing them remains remarkably consistent. The FTC is not waiting for new laws; it is using its existing Section 5 authority to aggressively root out “AI-powered frauds and scams.”[25] For businesses, the path forward requires a balance: Continue to innovate, but ensure your marketing team’s “hype” never outruns your engineering team’s “reality.”[26] In the age of AI, a truthful claim is still the best protection against a million-dollar consent decree.
[1] Kathleen Benway et al., Consumer Protection/FTC Advisory: The FTC Takes Aim at Deceptive AI Claims, Alston & Bird (Oct. 2, 2024) https://www.alston.com/en/insights/publications/2024/10/the-ftc-takes-aim-at-deceptive-ai-claims [https://perma.cc/F39Y-LT7R].
[2] Id; FTC Announces Crackdown on Deceptive AI Claims and Schemes, Federal Trade Commission (Sept. 25, 2024), https://www.ftc.gov/news-events/news/press-releases/2024/09/ftc-announces-crackdown-deceptive-ai-claims-schemes [https://perma.cc/5XJ3-7PB2].
[3] FTC Announces Crackdown on Deceptive AI Claims and Schemes, supra note 2.
[4] Id.
[5] Federal Trade Commission Act §5, Unfair or Deceptive Acts or Practices, at 1. https://www.federalreserve.gov/boarddocs/supmanual/cch/200806/ftca.pdf [https://perma.cc/DB7Z-AKWV]; Benway, supra note 2; FTC Announces Crackdown on Deceptive AI Claims and Schemes, supra note 4.
[6] Federal Trade Commission Act §5, supra note 5.
[7] Id. at 8.
[8] Id. at 1.
[9] Gombos Atila Robert, Truth in Advertising: New Standards for AI-Enhanced Local Ads, Jasmine Bus. Directory (June 12, 2025), https://www.jasminedirectory.com/blog/truth-in-advertising-new-standards-for-ai-enhanced-local-ads [https://perma.cc/VUN5-WXQ7].
[10] See David Ervin & Nicholas Pung, FTC Cracks Down on AI Model’s AI Detection Claims, Crowell & Moring (May 7, 2025), https://www.crowell.com/en/insights/client-alerts/ftc-cracks-down-on-ai-models-ai-detection-claims [https://perma.cc/REE7-R6HE]; See also Anthony E. DiResta, FTC Evaluating Deceptive Artificial Intelligence Claims, Holland & Knight (June 11, 2025), https://www.hklaw.com/en/insights/publications/2025/06/ftc-evaluating-deceptive-artificial-intelligence-claims [https://perma.cc/BPV2-8MT6].
[11] See Ervin & Pung, supra note 10; See also DiResta, supra note 10.
[12] FTC Order Requires Online Marketer to Pay $1 Million for Deceptive Claims that its AI Product Could Make Websites Compliant with Accessibility Guidelines, Federal Trade Commission (Jan. 3, 2025), https://www.ftc.gov/news-events/news/press-releases/2025/01/ftc-order-requires-online-marketer-pay-1-million-deceptive-claims-its-ai-product-could-make-websites [https://perma.cc/Z9EL-F39D].
[13] Id.
[14] Laura Kim et al., FTC Challenges Deceptive Artificial Intelligence Claims, Covington (June 4, 2025), https://www.globalpolicywatch.com/2025/06/ftc-challenges-deceptive-artificial-intelligence-claims [https://perma.cc/N69H-CCJN]; DiResta, supra note 11.
[15] Kim, supra note 14; DiResta, supra note 14.
[16] FTC Takes Action Against IntelliVision Technologies for Deceptive Claims About its Facial Recognition Software, Federal Trade Commission (Dec. 3, 2024), https://www.ftc.gov/news-events/news/press-releases/2024/12/ftc-takes-action-against-intellivision-technologies-deceptive-claims-about-its-facial-recognition [https://perma.cc/LQ6G-H5ER].
[17] Id; Benway, supra note 1.
[18] Workado claimed its AI detector was 98% accurate, but independent testing showed it was closer to 53%. See David Ervin & Nicholas Pung, FTC Cracks Down on AI Model’s AI Detection Claims, Crowell & Moring (May 7, 2025), https://www.crowell.com/en/insights/client-alerts/ftc-cracks-down-on-ai-models-ai-detection-claims [https://perma.cc/REE7-R6HE].
[19] Schemes like Ascend Ecom and Ecommerce Empire Builders lured consumers with promises of “passive income” through AI-powered storefronts that frequently failed to materialize, allegedly defrauding consumers of millions. See FTC Announces Crackdown on Deceptive AI Claims and Schemes, Federal Trade Commission (Sept. 25, 2024), https://www.ftc.gov/news-events/news/press-releases/2024/09/ftc-announces-crackdown-deceptive-ai-claims-schemes [https://perma.cc/5XJ3-7PB2].
[20] Benway, supra note 1.
[21] Anthony E. DiResta, FTC Evaluating Deceptive Artificial Intelligence Claims, Holland & Knight (June 11, 2025), https://www.hklaw.com/en/insights/publications/2025/06/ftc-evaluating-deceptive-artificial-intelligence-claims [https://perma.cc/BPV2-8MT6].
[22] Federal Trade Commission Act §5, Unfair or Deceptive Acts or Practices, at 4. https://www.federalreserve.gov/boarddocs/supmanual/cch/200806/ftca.pdf [https://perma.cc/DB7Z-AKWV].
[23] Gombos Atila Robert, Truth in Advertising: New Standards for AI-Enhanced Local Ads, Jasmine Bus. Directory (June 12, 2025), https://www.jasminedirectory.com/blog/truth-in-advertising-new-standards-for-ai-enhanced-local-ads [https://perma.cc/VUN5-WXQ7].
[24] FTC Takes Action Against IntelliVision Technologies for Deceptive Claims About its Facial Recognition Software, Federal Trade Commission (Dec. 3, 2024), https://www.ftc.gov/news-events/news/press-releases/2024/12/ftc-takes-action-against-intellivision-technologies-deceptive-claims-about-its-facial-recognition [https://perma.cc/LQ6G-H5ER].
[25] David Ervin & Nicholas Pung, FTC Cracks Down on AI Model’s AI Detection Claims, Crowell & Moring (May 7, 2025), https://www.crowell.com/en/insights/client-alerts/ftc-cracks-down-on-ai-models-ai-detection-claims [https://perma.cc/REE7-R6HE]; Alice S. Hrdy & Caitlin Zeytoonian, FTC Moves to Expand AI Deployment Oversight with ‘Operation AI Comply’, Morgan Lewis (Oct. 16, 2024), https://www.morganlewis.com/pubs/2024/10/ftc-moves-to-expand-ai-deployment-oversight-with-operation-ai-comply [https://perma.cc/T43F-KEE3].
[26] Hrdy & Zeytoonian, supra note 25.