Florida Deceptive and Unfair Trade Practices Act (FDUTPA)

The Florida Deceptive and Unfair Trade Practices Act, Fla. Stat. §§ 501.201–501.213, is the most versatile claim in Florida commercial litigation. Enacted in 1973 to give Florida its own “Little FTC Act,” it prohibits “unfair methods of competition, unconscionable acts or practices, and unfair or deceptive acts or practices in the conduct of any trade or commerce.” Unlike the Lanham Act, it requires no interstate commerce; unlike common-law fraud, it requires no proof of intent or of the plaintiff’s individual reliance; and unlike the FTC Act, it gives a private right of action, with attorney’s fees, to anyone aggrieved. It is pleaded in nearly every trademark, false-advertising, unfair-competition and consumer case filed in this state — and, since 2025, it has become the Florida Attorney General’s principal instrument against the developers and operators of artificial-intelligence systems and online platforms.

Our Florida intellectual property attorneys assert and defend FDUTPA claims in state and federal court, both alongside trademark, copyright and trade secret claims and as stand-alone actions. This summary explains the statute, its elements and remedies, its relationship to intellectual property enforcement, and — in detail — its application to emerging technologies, including generative AI, chatbots, algorithmic marketing, synthetic media and data practices. It closes with the key considerations for technology companies doing business in Florida.

What FDUTPA Is, and Where It Comes From

FDUTPA is modeled on Section 5 of the Federal Trade Commission Act and directs Florida courts to give “due consideration and great weight” to the FTC’s and federal courts’ interpretations of Section 5. Fla. Stat. § 501.204(2). Its operative prohibition, § 501.204(1), is deliberately broad, and the statute instructs that it be “construed liberally” to protect the consuming public and legitimate business enterprises from those who engage in unfair methods of competition or deceptive practices, § 501.202. “Trade or commerce” covers the advertising, soliciting, providing, offering or distributing of any good or service, property or thing of value, § 501.203(8), which Florida courts have read to reach essentially every commercial activity, including the operation of websites, applications and online platforms.

The statute is enforced two ways. The Department of Legal Affairs — the Attorney General — and state attorneys may investigate, issue subpoenas, and sue for injunctions, restitution and civil penalties, §§ 501.206–.2075. And “a person who has suffered a loss” as a result of a violation may sue for actual damages, and “anyone aggrieved” may sue for declaratory and injunctive relief, § 501.211. Both public and private enforcement have grown steadily; private FDUTPA claims are a fixture of Florida commercial litigation, and Attorney General enforcement has, in the last two years, been directed with unprecedented energy at technology companies.

Elements of the Cause of Action

Private claim for damages — § 501.211(2)

• A deceptive act or unfair practice in the conduct of trade or commerce;
• Causation — the practice caused the plaintiff’s loss, judged objectively by whether the practice was likely to deceive a consumer acting reasonably in the circumstances, rather than by the plaintiff’s subjective reliance; and
• Actual damages — measured as the difference between the market value of the product or service as delivered and its value as it should have been delivered, or the purchase price where the product is valueless.

These elements, drawn from Gavron v. Weather Shield Manufacturing, Inc., 819 F. Supp. 2d 1297 (S.D. Fla. 2011), and the Florida district courts of appeal, are the framework every Florida court applies. A plaintiff seeking only declaratory or injunctive relief under § 501.211(1) need not prove actual damages; it need only be “aggrieved” by the violation, a standard that includes competitors and other businesses harmed by a defendant’s practices.

Attorney General enforcement — §§ 501.207, 501.2075, 501.2077

The enforcing authority need prove neither causation nor damages to a particular consumer. It must show a violation of § 501.204 — a deceptive, unfair or unconscionable act or practice in trade or commerce — and may then obtain a declaratory judgment, an injunction, restitution or reimbursement for consumers, and civil penalties of up to $10,000 per violation, § 501.2075, rising to $15,000 per violation for practices victimizing senior citizens, military service members or persons with disabilities, § 501.2077. Each affected consumer and each communication can be a separate violation, which is why penalty exposure in platform cases is calculated in the millions.

Deceptive, Unfair and Unconscionable

The statute reaches three overlapping categories of conduct. A deceptive act is a representation, omission or practice that is likely to mislead a consumer acting reasonably in the circumstances, to the consumer’s detriment; the standard is probable, not merely possible, deception, and it covers omissions of material facts as well as affirmative statements. An unfair practice is one that offends established public policy or is immoral, unethical, oppressive, unscrupulous or substantially injurious to consumers — the formulation Florida courts borrowed from the FTC’s Sperry & Hutchinson standard — and, under the FTC’s modern unfairness test that Florida courts also apply, one that causes substantial injury not reasonably avoidable by consumers and not outweighed by countervailing benefits. An unconscionable act, less frequently litigated, requires both a procedural component (an absence of meaningful choice) and a substantive component (terms unreasonably favorable to the other party). Design choices, default settings and the architecture of an online service can be “practices” under any of the three headings, which is the doctrinal basis on which the Attorney General’s technology cases proceed.

Per Se Violations: Borrowed Statutes and FTC Rules

Section 501.203(3) provides that a violation of FDUTPA may be based on “[a]ny rules promulgated pursuant to the Federal Trade Commission Act” or “[a]ny law, statute, rule, regulation, or ordinance which proscribes unfair methods of competition, or unfair, deceptive, or unconscionable acts or practices.” This provision converts a growing list of federal and state requirements into FDUTPA violations, with FDUTPA’s remedies attached. The following are the borrowed sources most often invoked in technology and intellectual-property matters.

Remedies Compared

Two limits deserve emphasis. First, FDUTPA does not provide consequential damages: the Eleventh Circuit confirmed in Deltona Transformer Corp. v. The NOCO Co. (Aug. 2026) that lost goodwill and reputation are consequential, not actual, damages, so a trademark plaintiff that proves a FDUTPA violation through its Lanham Act claim is entitled to an injunction but must look to the Lanham Act for money. Second, the fee provision is two-edged: a prevailing defendant may recover fees, and courts weigh the merit of the claim and the parties’ conduct in deciding whether to award them, which makes a weak FDUTPA count a liability rather than a free addition to a complaint.

FDUTPA and Intellectual Property Enforcement

Because a Lanham Act violation is a per se FDUTPA violation, our attorneys plead FDUTPA in trademark, false-advertising and false-endorsement cases as a matter of course. The FDUTPA count adds a Florida statutory claim that does not depend on interstate commerce, supports declaratory and injunctive relief for competitors as “aggrieved” persons, and provides a fee-shifting provision independent of the Lanham Act’s “exceptional case” standard. It also reaches conduct the Lanham Act does not: misleading omissions, unfair business methods that do not involve a mark, and practices that injure consumers without confusing them. Three cautions govern its use. The damages count must rest on a difference-in-value or purchase-price theory, not on goodwill; a FDUTPA count that “sounds in fraud” — that alleges the defendant made specific false statements — must be pleaded with Rule 9(b) particularity in federal court, as the Eleventh Circuit held in Pop v. LuliFama.com LLC (2025); and a competitor plaintiff should plead its status as an aggrieved person and the consumer-facing nature of the practice, since some Florida courts require that the challenged practice injure consumers, not merely a rival. See our Trademark Matters and Litigation & Enforcement summaries.

FDUTPA and Emerging Technologies

FDUTPA was written for a world of door-to-door sales and print advertising, but its breadth has made it the principal Florida law governing the commercial conduct of technology companies — and, in the last two years, the statute on which the State has built an aggressive enforcement program against artificial-intelligence developers and online platforms. The following sections describe how the statute is being applied and where the exposure lies.

Generative AI: State of Florida v. OpenAI

On June 1, 2026, the Attorney General filed suit in the Circuit Court for Highlands County against OpenAI and its chief executive, asserting two FDUTPA counts and a public-nuisance claim. Count I alleges unfair, immoral and deceptive practices: marketing ChatGPT as safe and reliable (“Safety ... core to our mission”) while knowing that it produces fabricated information and can provide dangerous assistance; failing to warn of those dangers; designing features that encourage compulsive use; deploying anthropomorphic features that the complaint characterizes as manipulating engagement; and releasing the product to minors without adequate age verification or parental controls. Count II alleges that the same conduct is unconscionable. The complaint invokes the FTC’s COPPA Rule as a per se violation for the collection of data from users under 13, and it cites the April 2025 Florida State University shooting — in which, the State alleges, the gunman consulted ChatGPT for tactical information — as evidence that the product “proactively aids, abets, and promotes dangerous activities.” The State seeks disgorgement, restitution, civil penalties of up to $10,000 per violation, and a permanent injunction, and demands a jury. The filing followed a criminal investigation, announced April 21, 2026, in which the Attorney General subpoenaed OpenAI’s safety policies, training materials and organizational records on an aiding-and-abetting theory, and it was followed in September 2026 by the Attorney General’s announcement of proposed legislation that would impose criminal sanctions, fines, restitution, court-ordered monitorship and suspension of Florida operations on companies with “practical control” over AI systems whose chatbots facilitate crimes, self-harm, child exploitation or unlicensed medical advice.

Whatever its ultimate outcome, the case establishes the template the State will apply to any generative-AI product offered to Floridians. Its theories do not depend on any AI-specific statute; they rest on the ordinary FDUTPA propositions that a safety representation must be true, that a known risk must be disclosed, that a product design that exploits users is an unfair practice, and that a service must not collect children’s data without parental consent. Every developer and deployer of an AI system that interacts with Florida consumers should assume those propositions will be applied to it.

AI chatbots, companions and minors

The OpenAI action is one of a series. In 2025 and 2026 the Attorney General sued Snap and then TikTok and ByteDance under FDUTPA and § 501.1736, and has investigated or sued other platforms, including Roblox and Netflix, over children’s data and safety, alleging that platform features are designed to be addictive to minors, that age-verification and parental-consent mechanisms are inadequate, and that the companies’ public representations about safety are false. The TikTok complaint, filed June 15, 2026 in St. Lucie County, seeks civil penalties of up to $50,000 per violation, punitive damages and disgorgement. AI “companion” and character chatbots that engage minors in emotionally dependent relationships, provide sexual content, or discuss self-harm are the next obvious targets, and the Attorney General’s September 2026 proposal expressly addresses chatbots that encourage self-harm and generate child sexual abuse material. Companies that offer conversational AI to Florida users should treat age assurance, content safeguards, crisis-response protocols, parental controls and the accuracy of every safety claim as FDUTPA compliance obligations.

“AI-washing” and performance claims

Claims that a product uses artificial intelligence, that an AI system performs at a stated level of accuracy, that it is “unbiased,” “hallucination-free,” “clinically validated” or “as good as a lawyer,” or that it can accomplish a specified result are advertising claims that must be truthful and substantiated when made. The FTC’s 2024 “Operation AI Comply” actions against companies that oversold AI capabilities are the federal model, and § 501.204(2) directs Florida courts to follow them. A Florida business that markets AI features it does not have, or that overstates what its models do, is exposed to competitor suits for injunctive relief and fees, consumer class actions for the price premium paid, and Attorney General penalties.

AI-generated reviews, testimonials, endorsements and personas

The FTC’s Rule on Consumer Reviews and Testimonials, effective October 2024, prohibits creating, buying or disseminating fake reviews and testimonials — expressly including those generated by AI — as well as reviews by insiders without disclosure and the suppression of negative reviews. Because it is an FTC trade regulation rule, its violation is a per se FDUTPA violation under § 501.203(3)(a). The FTC’s Endorsement Guides likewise require disclosure of material connections, and their principles apply to virtual influencers and AI-generated spokespersons: a synthetic persona presenting itself as a satisfied customer, or an AI-generated “expert” endorsement, is a deceptive practice. Florida businesses that use AI to generate marketing copy should ensure that no output is presented as a genuine consumer or expert statement, that influencer relationships — human or synthetic — are disclosed, and that review-management tools do not filter by sentiment.

Synthetic media, deepfakes and digital replicas

AI-generated depictions of real people raise FDUTPA exposure in several ways. A synthetic image, voice or video of a celebrity, executive or ordinary person used to sell a product is a false endorsement under the Lanham Act and therefore a per se FDUTPA violation, in addition to violating Florida’s right of publicity, § 540.08. A platform that fails to remove nonconsensual altered sexual depictions within 48 hours of notice violates Brooke’s Law and, by its terms, FDUTPA. Political advertisements that use generative AI to depict a real person doing something that did not occur require a disclaimer under § 106.145. And a business that uses a digital replica of a real person without disclosing that the depiction is synthetic, in a context where consumers would be misled, engages in a deceptive practice regardless of whether any specific statute applies. Our earlier commentary on the Copyright Office’s digital-replica report and the pending federal NO FAKES Act addresses the licensing and consent practices that avoid these claims.

Data collection, training data and privacy representations

Every AI system is built on data, and the collection, use and disclosure of Florida consumers’ data is regulated by an overlapping set of provisions that FDUTPA enforces. The Florida Digital Bill of Rights imposes obligations on large controllers and specific restrictions on the sale of sensitive data and on the collection of data from children; the Florida Information Protection Act requires reasonable security and breach notification; the COPPA Rule governs children under 13; and the FTC’s Section 5 jurisprudence — which Florida courts follow — treats a privacy policy that does not match actual practice as deceptive and the use of consumer data to train models without disclosure as potentially unfair. The FTC has, in several matters, ordered the deletion of models trained on improperly collected data (“algorithmic disgorgement”), a remedy available in principle under FDUTPA’s equitable powers. Companies that train or fine-tune models on customer data, that use conversational data for product improvement, or that share data with AI vendors should ensure that their privacy policies and consent flows describe those uses accurately.

Dark patterns, subscriptions and algorithmic design

Interface designs that steer consumers into purchases, subscriptions or data disclosures they did not intend — pre-checked boxes, obstructed cancellation, confirmshaming, drip pricing, disguised advertisements — are unfair or deceptive practices under the FTC’s guidance that Florida courts follow, and auto-renewal practices are separately governed by § 501.165. As AI systems increasingly generate and personalize interfaces, offers and prices in real time, the same principles apply to the outputs: an algorithm that presents different consumers with different prices or terms on the basis of protected characteristics or inferred vulnerability, or that generates persuasive content designed to exploit a consumer’s circumstances, is a practice that can be challenged as unfair whether or not a human designed each screen.

Emerging-technology exposure at a glance

Defenses and Limits

FDUTPA has meaningful boundaries, and a defendant’s first task is to identify which apply. The statute does not apply to an act or practice “required or specifically permitted by federal or state law,” § 501.212(1), which shields conduct compelled by regulation but not conduct merely unaddressed by it. It does not apply to claims for personal injury or death or for damage to property other than the property that is the subject of the consumer transaction, § 501.212(3), which is why the Attorney General’s technology complaints pair FDUTPA with public-nuisance and other tort theories. It exempts certain regulated entities and activities, including some banking, insurance and securities conduct, § 501.212(4)–(6). Its private damages remedy is limited to actual damages as defined above, excluding consequential loss, and the four-year limitations period runs from the violation. In federal court, a FDUTPA claim grounded in specific misrepresentations must satisfy Rule 9(b). Puffery — general, non-measurable claims of superiority — is not actionable, though the line between puffery and a measurable AI-performance claim is narrow. And a defendant that prevails may seek its fees under § 501.2105, which gives every FDUTPA plaintiff a reason to plead carefully. Our Litigation Defense summary describes our defense practice.

Key Considerations for Technology Companies

• Every product claim is an advertising claim. Statements about what an AI system does, how safe or accurate it is, and what safeguards it has are representations that must be true and substantiated when made. Review marketing copy, app-store descriptions, model cards and terms of service against actual performance, and retain the substantiation.
• Disclose known risks. Hallucination, bias, failure modes and the limits of safety filters are material facts. The State’s theory in the OpenAI case is that omitting them is deceptive; a clear, prominent disclosure is the defense.
• Treat minors as a separate compliance problem. Age assurance, parental consent and controls, content restrictions and § 501.1736’s requirements apply to any service minors can reach. The penalty exposure for children’s data and minors’ accounts is five times the general FDUTPA maximum.
• Audit engagement design. Features intended to maximize time on service, streaks, autoplay, variable rewards and anthropomorphic companionship are being litigated as unfair practices. Document the purpose and the safeguards for each.
• Keep synthetic content honest. No AI-generated review, testimonial, endorsement or persona should be presented as a real person’s statement; every material connection should be disclosed; and every depiction of a real person should be licensed and, where consumers could be misled, labeled.
• Match the privacy policy to the data pipeline. If customer or conversational data is used to train or improve models or is shared with AI vendors, the policy must say so and the consent must cover it. Where a data-deletion request is honored, consider whether model outputs are affected.
• Design cancellation and consent flows for the consumer, not the funnel. Auto-renewal disclosures, one-click cancellation, unchecked boxes and honest pricing are inexpensive; dark-pattern claims are not.
• Expect the Attorney General. The Department of Legal Affairs has made technology enforcement a priority, uses investigative subpoenas aggressively, and pleads FDUTPA with public-nuisance and statutory theories that carry penalties of $10,000 to $50,000 per violation. A company that receives a civil investigative demand should respond with counsel from the first day.
• Use FDUTPA offensively. Competitors that misrepresent their AI capabilities, use fake reviews, or trade on your brand are FDUTPA defendants. A competitor is an “aggrieved” person entitled to injunctive relief and fees, and a Lanham Act violation supplies the FDUTPA claim automatically. See our Cease and Desist Letters summary.
• Plead and defend with precision. Damages counts must rest on a difference-in-value theory; fraud-based counts must satisfy Rule 9(b); competitor claims should articulate consumer injury; and every FDUTPA count should be evaluated against the two-way fee provision before it is filed.

Full Capability

Our FDUTPA Services Include