Browse by topic
12 themes · across 61 jurisdictionsThe same obligations, sliced by what they regulate rather than where. Each theme gathers every tracked rule on it — from every jurisdiction — onto one primary-sourced page.
| Topic | Obligations | Jurisdictions | Next deadline |
|---|---|---|---|
| Deepfakes & content labellingWhere AI-generated images, video, audio and text must be disclosed, watermarked or removed — the binding rules and the proposals. | 25 | 12 | 1 Oct 2026 |
| Transparency & disclosureWhen people have to be told they are dealing with AI, and content has to be marked as model-generated — chatbot disclosure, content labelling and training-data transparency. | 99 | 57 | 17 Sep 2026 |
| High-risk AI systemsHow the major risk-based regimes define high-risk or high-impact AI, and the assessments, oversight and documentation that classification triggers. | 43 | 36 | 1 Jan 2027 |
| Frontier & general-purpose AIThe rules aimed at foundation-model and frontier developers — safety frameworks, incident reporting and model documentation. | 13 | 6 | 1 Oct 2026 |
| Copyright & intellectual propertyWhere copyright law restricts — or expressly permits — the use of protected works to build and run AI systems, and what rightsholders can reserve. | 7 | 5 | 1 Mar 2027 |
| Training data & TDMRules on where model training data may come from and what has to be said about it — text-and-data-mining permissions, scraping limits and public training-content summaries. | 7 | 5 | 1 Mar 2027 |
| Prohibited AI practicesThe AI uses that are forbidden outright — manipulative systems, social scoring, non-consensual intimate imagery and algorithmic price coordination — regardless of safeguards. | 25 | 10 | 1 Dec 2026 |
| AI in hiring & employmentThe rules governing automated hiring and employment decisions — bias audits, candidate notice and a right to contest. | 45 | 41 | 1 Jan 2027 |
| Companion & conversational AIThe rules aimed at consumer-facing chatbots and AI companions — telling users they are talking to a machine, crisis-referral protocols, and safeguards for younger users. | 13 | 2 | 1 Jan 2027 |
| AI therapy & mental healthWhere AI is barred from practising psychotherapy, presenting itself as a licensed clinician, or being advertised as mental-health care — and where a licensed human has to stay in the loop. | 14 | 1 | 1 Jan 2027 |
| Minors & age-appropriate AIThe provisions written specifically for users under 18 — age signals and verification, parental controls, training-data limits and design constraints on AI systems children can reach. | 15 | 4 | 17 Sep 2026 |
| Healthcare & coverage decisionsThe rules on AI in medical-necessity, prior-authorization, utilization-review and claims decisions — where a licensed clinician has to make the call and what an insurer has to disclose. | 15 | 1 | 1 Oct 2026 |
Not legal advice. A theme gathers the obligations we tag to it; an obligation can sit under more than one theme. See each row's source for the legal text.