Autonomy is a person’s capacity for self‑rule: to form, revise, and act on values and reasons without deception or coercion.
In AI settings, autonomy is supported when systems are understandable, options are presented fairly, and consent is meaningful. It is threatened by dark patterns, opaque recommender systems, and designs that exploit vulnerabilities or hide tradeoffs. Safeguarding autonomy calls for transparency people can actually use, privacy by default, clear opt‑out paths, and governance that prevents manipulative or deceptive practices.
Disclaimer: Our global network of contributors to the AI & Human Rights Index is currently writing these articles and glossary entries. This particular page is currently in the recruitment and research stage. Please return later to see where this page is in the editorial workflow. Thank you! We look forward to learning with and from you.