EU political agreement on the Digital Omnibus (AI omnibus) — confirmed high-risk AI timeline

Tag: S-2026-05-07-eu-ai-omnibus-agreement Type: report (European Commission policy page / press communication) Author(s): European Commission (AI Act policy page; AI Office) Date of source: 2026-05-07 (date political agreement reached; reported on the Commission AI Act page, last updated 11 May 2026) Date ingested: 2026-06-01 Authority weight: high — primary EU institutional source (European Commission). Raw file: S-2026-05-07-eu-ai-omnibus-agreement.md. External URL: https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai and https://digital-strategy.ec.europa.eu/en/news/eu-agrees-simplify-ai-rules-boost-innovation-and-ban-nudification-apps-protect-citizens

What it claims

The Commission’s AI Act page states that the Digital Package on Simplification — the legislative proposal dubbed the “AI omnibus” — was adopted by the Commission on 19 November 2025 and that a political agreement was reached on 7 May 2026. Following the political agreement, a clear implementation timeline is set for the rules governing high-risk AI systems:

• Rules for systems used in certain high-risk areas (biometrics, critical infrastructure, education, employment, migration, asylum and border control) will apply from 2 December 2027.
• Rules for high-risk systems integrated into products (e.g. lifts, toys) will apply from 2 August 2028.

The page frames this as ensuring the rules apply “when companies have the right support tools to facilitate implementation, such as standards.” Other agreed amendments include: prohibiting AI systems that generate non-consensual sexually explicit content or CSAM (e.g. “nudification” apps); reinforcing the AI Office’s powers and centralising oversight of systems built on GPAI models; extending simplified requirements (e.g. simplified technical documentation) to SMEs and small mid-caps; widening access to regulatory sandboxes including an EU-level sandbox; and clarifying the interplay between the AI Act and EU product-safety law (notably the Machinery Regulation). The overall AI Act still entered into force 1 August 2024 and remains otherwise applicable from 2 August 2026, with GPAI obligations applicable since 2 August 2025.

Notable quotes

“This legislative proposal (dubbed as the ‘AI omnibus’) has been adopted on 19 November 2025 and a political agreement was reached on 7 May 2026.” — Commission AI Act page (last update 11 May 2026)

“Rules for systems used in certain high-risk areas — including biometrics, critical infrastructure, education, employment, migration, asylum and border control — will apply from 2 December 2027.” — Commission AI Act page

What’s speculative vs. asserted

• Asserted: that a political agreement was reached on 7 May 2026; the 2 December 2027 (high-risk areas) and 2 August 2028 (product-embedded) application dates; the listed amendments; that GPAI and Article 50 transparency tracks are unaffected (transparency rules still come into effect August 2026).
• Speculative / not yet confirmed: the date of formal adoption / entry into force of the amended text. A political agreement between co-legislators precedes formal adoption; the page presents the new dates as the agreed outcome but does not state that the amending regulation has been formally adopted and published in the Official Journal.

Topics this feeds

• EU AI Act — supersedes the prior “delay is proposed but not yet law” framing for the high-risk timeline.
• EU AI Office — the agreement reinforces the AI Office’s powers and centralises GPAI-based-system oversight.

Open questions raised

• When the amended AI Act text will be formally adopted and published (political agreement ≠ enacted law).
• Whether harmonised standards (CEN-CENELEC JTC 21) will be ready ahead of the 2 December 2027 high-risk application date.
• How national competent authorities will sequence supervision between the August 2026 (GPAI / transparency) and December 2027 (high-risk areas) tracks.

Ingestion note

Captured by direct WebFetch of the Commission AI Act policy page (https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), last updated 11 May 2026. This source updates S-2025-11-19-eu-digital-omnibus (the 19 November 2025 proposal), which the wiki had recorded as “not yet law”; the 7 May 2026 political agreement is a material development but remains a political agreement pending formal adoption — preserve that nuance.