AI Policy Registry
Significant AI policy actions across the US, EU, UK, China, and international forums. Editorial, structured, source-linked.
The other “AI policy tracker” resources online tend to be either expert paywalls or wikis that go stale. This is the third option: a structured editorial registry, published as a free JSON API for agents, with each entry linking to the canonical government source.
Status semantics are explicit: active /phased /pending /rescinded /vetoed. Scope tags (transparency, safety, high-risk, deepfakes, export-controls, etc.) make filtering obvious.
Free agent endpoint
/api/policy/ai/registryFull registry. Filters:?jurisdiction=&type=&status=&scope=. Each entry returns id, title, jurisdiction, type, status, enacted_date, effective_date, summary, source_url, citations, scope[].
Frequently asked questions
- What does this catalog cover?
- Significant AI policy actions across six jurisdictions: US Federal (executive orders, BIS export controls), US State (Colorado SB 24-205, Utah HB 333, California AB 2013 / SB 942 / SB 1047, etc.), EU (AI Act 2024/1689 phased rollout), UK (pro-innovation paper, AI Safety Institute), China (CAC GenAI and Deep Synthesis Provisions), and international declarations (Bletchley, Seoul). Each entry carries jurisdiction, type, status, enacted date, effective date, scope tags, and a link to the canonical source.
- Who curates the registry?
- TensorFeed editorial. The underlying acts, orders, and regulations are government publications (US federal works are public domain; EU and UK government works are reproduced under their open-government terms with citation; China CAC publications are official rulemaking texts). Summaries are our own; we link directly to the canonical government publication for each entry so you can verify the source yourself.
- How fresh is it?
- Editorial. Updates land on redeploy when significant new policy is enacted or when status changes (active to rescinded, pending to active, etc). The registry is small enough to maintain accurately by hand; we'd rather ship 16 well-curated entries than 100 with stale status fields.
- Can I use this commercially?
- Yes. Underlying government publications are free to redistribute under their respective public-records terms. The TensorFeed registry shape (titles, summaries, scope tags) is editorial work product and we encourage agents and analysts to consume it via the JSON API. Each entry preserves the canonical source URL so end-users can verify.