Author: Nita Farahany Source: https://nitafarahany.substack.com/p/20-clicks-to-cancel Published: February 2, 2026
"20 Clicks to Cancel" is a February 2, 2026 essay by Nita Farahany, the third installment in her published account of an Advanced Topics in AI Law and Policy course (Class 3 of roughly 30). It opens the Week 3 module on the Architecture of Choice and dark patterns, using the Federal Trade Commission's $25 million Amazon Prime settlement as its anchor case. The essay reframes "user choice" from a static legal question of whether a consumer consented into a question of the conditions under which they consented, and previews the argument that a disclosure-based consumer-protection toolkit is structurally unable to reach the design of choice environments.
Summary of argument
The essay's organizing claim is that the design of a choice environment matters as much as the choices it presents. It treats consent not as a binary fact established at the moment a user agrees but as a product of the design conditions surrounding that agreement, and argues that existing consumer-protection law, built around disclosure, is poorly equipped to address manipulation that operates through design rather than through false statements.
Key claims
The Amazon Prime cancellation flow serves as the canonical dark pattern. Signing up for Prime required one click, while cancellation required navigating a multi-screen process the FTC named the "Iliad Flow," after Homer's epic about an interminable war. The FTC's 2023 enforcement action and $25 million settlement (FTC v. Amazon, Inc.) treated this asymmetry as a deceptive practice under FTC Act §5, even though every required disclosure was technically present.
Dark patterns are distinguished from fraud. Because the disclosure is present and the consumer technically agreed, dark patterns generally do not satisfy the misrepresentation element of traditional fraud. The essay locates their harm instead in the engineering of friction between the consumer and an exit. The FTC's authority under the §5 "unfairness" prong is described as the closest US doctrine to a remedy, but one that requires showing "substantial injury not reasonably avoidable," itself a contested standard as applied to design choices.
The essay draws on the Mathur et al. (2019) taxonomy of dark patterns, a Princeton study of roughly 11,000 e-commerce sites that found 1,818 instances of dark patterns across 1,254 sites. Its categories include sneak-into-basket, hidden costs, forced continuity, misdirection, urgency manipulation, scarcity claims, social proof manipulation, and confirm-shaming. Farahany characterizes this empirical base as substantial enough to support regulatory specificity.
A central point is the limit of disclosure theory. Most consumer-protection law assumes the problem is information asymmetry, on the premise that telling users more lets them choose better. Dark patterns invert that assumption: users can know they are being manipulated and remain manipulated, because the design exploits cognitive load and friction below conscious awareness. On this account, disclosure is structurally inadequate to the problem.
The essay previews a regulatory shift from disclosure toward architecture. California's AB 587, the EU's Digital Services Act Article 25, and the FTC's "click-to-cancel" rule are each described as attempts to mandate specific design properties rather than additional disclosures. Whether such mandates survive First Amendment and commercial-speech scrutiny is left as an open question that Week 9 of the course returns to.
Relationships
- part-of: Nita Farahany Advanced Topics course (Class 3 of ~30)
- introduces: dark-patterns governance framing (planned concept)
- related: Cognitive Liberty
- previous: Inside My Advanced Topics Class 2.3: The Laws That Miss the Point (Farahany, January 2026) next: Inside My Advanced Topics Class 3.2: Why Dark Patterns Work (Farahany, February 2026)