Cognitive liberty is the right to self-determination over the substrate of decision-making itself, comprising mental privacy, freedom of thought, and the capacity to direct one's own cognition free of covert influence. It is the organizing concept of Nita Farahany's scholarship and the through-line of her Advanced Topics in AI Law and Policy course (Spring 2026).
Component rights
The concept comprises three component rights:
- Mental privacy — the right to keep one's thoughts, intentions, and inferred mental states from being read, recorded, or inferred without consent.
- Freedom of thought — the right to think without manipulation by systems that exploit cognitive vulnerabilities below conscious awareness.
- Self-determination in cognition — the right to direct one's own attention, reasoning, and decision-making free of architectural coercion.
Background
Farahany's account holds that existing US legal vocabularies each capture a slice of the territory but not the whole (per Class 1.3):
- Informed consent (medical/research) assumes a discrete decision moment and a defined intervention. It is built for one-off transactions, not continuous environmental shaping.
- Capacity (contracts/wills) is a binary status determined at the moment of decision. It does not capture gradations or population-level erosion.
- Undue influence (estates/elder law) is the closest analog, but requires identifying a specific influencing actor with a specific motive.
- Competency (criminal procedure) protects against state coercion of testimony only at the trial stage and only against the government.
None of these was designed for ambient, multi-actor, continuous influence at platform scale, which is the gap cognitive liberty is meant to address.
Doctrinal implications
Cognitive liberty is framed to supplement, not replace, existing vocabularies. Its operational consequence is that the law should treat the substrate of decision-making — attention, mood, mental state, inferred personality, neural signal — as a protectable interest in itself, not just as an instrumentality to other interests.
The Advanced Topics course works through concrete doctrinal applications:
- Attention (Class 2) — recognizing a "right to direct one's own attention" alongside the existing right to access information.
- Dark patterns (Class 3) — moving consumer-protection law from disclosure-theory to architecture-theory.
- AI companions (Class 5) — recognizing harm from fabricated rapport even where users "know" they are talking to AI.
- AI personality assessment (Class 6) — protecting against inferred-portrait harm where the inference is made and used without consent.
- Fifth Amendment (Class 10) — extending the Hubbell mental-labor doctrine to AI conversation logs and neural signals.
- First Amendment (Class 11) — recognizing user cognitive autonomy as an interest the doctrine has so far refused to weigh against platform editorial discretion.
International reception
Several jurisdictions and intergovernmental bodies have recognized related rights:
- Chile adopted a 2021 constitutional reform: an explicit "neurorights" amendment recognizing mental privacy and cognitive integrity.
- UNESCO issued a Recommendation on the Ethics of Neurotechnology (2025) treating mental privacy and cognitive liberty as foundational principles.
- OECD issued a Recommendation on Responsible Innovation in Neurotechnology (2019).
The US has no analogous federal recognition; cognitive-liberty arguments operate as commentary on existing doctrine rather than as standalone doctrinal claims.
Professional-body recognition
The mental-privacy component has been asserted outside the neurotechnology-specific instruments above. The American Psychological Association's June 2025 health advisory on AI and adolescent well-being asks AI platforms to recognize that data they collect, "including biometric and neural information from emerging technologies, can provide insights into mental states and cognitive processes," and states that systems "must safeguard this sensitive information and uphold adolescents' basic right to privacy" (Artificial Intelligence and Adolescent Well-being: An APA Health Advisory (June 2025)). The framing matters for the concept's reach: the claim is made about ordinary AI products collecting adolescent data, not about dedicated neurotechnology, which extends the mental-privacy argument from neural interfaces to inference from conventional behavioral and biometric data. The same advisory asks platforms to limit use of adolescent data for "personalized marketing that exploits their immature brain development," which is a freedom-from-manipulation claim rather than a privacy claim.
See also
- Three Privacy Problems AI Creates
- Two Conditions and Three Weapons of AI Manipulation
- Farahany Cognitive Evidence Spectrum (4 Categories)
Relationships
- instance-of: AI Governance (umbrella)
- depends-on: Nita Farahany
- supports: Inside My Advanced Topics in AI Law and Policy Class 1: From How AI Works to What AI Does (Farahany, January 2026), Inside My Advanced Topics Class 1.3: Protecting Autonomy in Law, Take 1 (Farahany, January 2026), Inside My Advanced Topics Class 10.1: The Government Can Take Your Blood. Can It Take Your Thoughts? (Farahany, March 2026), Inside My Advanced Topics Class 10.3: When the Interface Is Neural (Farahany, April 2026)
- related: Three Privacy Problems AI Creates, Two Conditions and Three Weapons of AI Manipulation, Farahany Cognitive Evidence Spectrum (4 Categories)
- supports: Artificial Intelligence and Adolescent Well-being: An APA Health Advisory (June 2025) — asserts mental privacy over biometric and neural data collected by ordinary AI products
- related: American Psychological Association (APA)