An experiment's report page: a tested kernel scores committed token counts (measured with Anthropic's count_tokens endpoint by a local script) and renders the findings. The page runs no model and collects nothing.
Runs on the owner's own machines · AI: none · map as of 2026-09-30 · law checked 2026-09-30
Not engaged
No AI system (Art 3(1)): the build runs no model that infers its outputs, so the AI Act places no duty on it.
Not engaged
No personal data is processed, so no GDPR / UK GDPR duty arises from the build.
0 apply · 0 in force
Nothing further scheduled.
| From | What applies | Where it says so |
|---|---|---|
| 2025-02-02 | AI Act Chapters I and II apply — including Article 4 (AI literacy) and the Article 5 prohibitions | Art 113, third para, point (a) |
| 2026-02-05 | UK: Data (Use and Access) Act 2025 data protection provisions in force — including section 80 (new Articles 22A–22D on automated decision-making) and section 85 / Schedule 7 (international transfers) | SI 2026/82, reg 2(j), (m), (z9) |
| 2026-07-27 | Digital Omnibus on AI (Regulation (EU) 2026/1744) in force | Reg 2026/1744, Art 4 |
| 2026-08-02 | AI Act general date of application — including Article 50 transparency | Art 113, second para |
| 2026-12-02 | AI Act Article 50(2) marking for generative systems already on the market before 2 August 2026; and the new Article 5(1) points (ba) and (bb) | Art 111(4) and Art 113, third para, point (a), as amended by Reg 2026/1744 |
| 2027-12-02 | AI Act high-risk obligations (Chapter III, Sections 1–3) for Annex III systems — moved from 2 August 2026 by the Digital Omnibus | Art 113, third para, point (c)(i), as replaced by Reg 2026/1744 |
| 2028-08-02 | AI Act high-risk obligations for Annex I (product-safety) systems | Art 113, third para, point (c)(ii), as replaced by Reg 2026/1744 |