European AI Office policy page — AI Omnibus entered into force 27 July 2026 (page last updated 13 August 2026)

Tag: S-2026-08-13-ec-ai-office-page-omnibus-in-force Type: article (European Commission policy web page, “European AI Office”, digital-strategy.ec.europa.eu) Author(s): European Commission, DG CONNECT (European AI Office) Date of source: 2026-08-13 (page “Last update”; the entry-into-force event it records is 27 July 2026) Date ingested: 2026-09-08 (retrieved during the 29 Aug–7 Sep 2026 catch-up scan; outside the window — ingested as a primary-source verification only, no Open Brain capture) Authority weight: high — canonical Commission page for the AI Office, retrieved in full, linking directly to the Official Journal reference (OJ L 2026/1744); the OJ text itself was not fetched. Raw file: S-2026-08-13-ec-ai-office-page-omnibus-in-force.md. Page: digital-strategy.ec.europa.eu/en/policies/ai-office.

What it claims

The Commission’s European AI Office page states that “In June 2026, the ‘AI omnibus’ (part of the Digital Simplification Package) was adopted, introducing targeted amendments to the AI Act to ensure the rules remain clear, simple, and innovation-friendly. The amendments entered into force on 27 July”, hyperlinking the entry-into-force statement to the Official Journal reference OJ L 2026/1744. The site’s news listing carries a matching 27 July 2026 item, “AI Omnibus enters into force”.

The page also describes the AI Office’s mandate and structure — more than 125 staff in six units (Excellence in AI and Robotics; Regulation and Compliance; AI Safety; AI Innovation and Policy Coordination; AI for Societal Good; AI in Health and Life Science) plus a Lead Scientific Adviser and an International Affairs Adviser — and its AI Act tasks: supporting coherent application across Member States, developing evaluation tools and benchmarks for GPAI models and classifying models with systemic risk, drawing up codes of practice, preparing guidance and implementing/delegated acts, investigating possible infringements (inspections, documentation requests, model access), and requesting corrective action, restricting model availability and issuing fines. It records that the AI Office is recruiting around 40 contractual agents to strengthen its enforcement team (technology specialists, legal officers, operations specialists, paralegals), applications closing 8 September 2026.

Notable quotes

• “The amendments entered into force on 27 July.” (section “Initiatives promoting the development and adoption of AI”)
• “Currently, the AI Office is recruiting around 40 contractual agents to strengthen its enforcement team.” (section “Job opportunities and collaboration”)

What’s speculative vs. asserted

• Asserted (as fact): adoption of the AI Omnibus in June 2026 and entry into force on 27 July 2026, with the OJ L 2026/1744 reference; the AI Office’s structure, headcount and task list; the enforcement-team recruitment and its 8 Sep 2026 deadline.
• Not stated: the identity of any “two new prohibitions” applying from 2 December 2026 (a claim in the Credo AI playbook S-2026-08-25-credo-eu-omnibus-playbook); the OJ text was not retrieved to check it.
• Ingesting-agent note: the page’s “June 2026” adoption wording is consistent with the Council’s final adoption on 29 June 2026 S-2026-06-29-council-ai-omnibus-final-adoption.

Topics this feeds

• EU AI Act — resolves (on primary footing) the Official-Journal / entry-into-force half of the open question raised by the vendor-relayed 27 July 2026 date; the “two new prohibitions” half remains open. Also supplies an enforcement-capacity datapoint (≈40 additional enforcement staff being recruited).

Open questions raised

• What does OJ L 2026/1744 say about prohibitions applying from 2 December 2026 (the vendor-asserted “two new prohibitions”) — is this the CSAM/nudification ban already recorded on EU AI Act plus one other, or a mischaracterisation?