EC high-risk classification guidelines — consultation extension to 23 July 2026 (primary page, directly retrieved)
Tag: S-2026-07-01-ec-highrisk-consultation-extension-confirmed Type: report (primary Commission consultation page, directly retrieved during the daily scan) Author(s): European Commission — DG CONNECT (AI Office) Date of source: 2026-06-16 (the consultation page’s own “Last update”; opening 19 May 2026, closing 23 July 2026) Date ingested: 2026-07-01 Authority weight: high — this is the primary Commission consultation page for the high-risk classification guidelines, directly retrieved (not a secondary tracker). It settles the deadline question the wiki had been carrying as unresolved. Raw file: S-2026-07-01-ec-highrisk-consultation-extension-confirmed.md. External URL: https://digital-strategy.ec.europa.eu/en/consultations/targeted-consultation-draft-guidelines-classification-high-risk-artificial-intelligence-systems
What it claims
The Commission’s targeted consultation on the draft guidelines for the classification of high-risk AI systems (Article 6 EU AI Act; draft published 19 May 2026) was originally open for six weeks, until 23 June 2026. The page states directly: “the Commission has received requests from several associations and other stakeholders to extend it further by 4 weeks… the deadline was extended to 23 July 2026.” The final guidelines will be adopted by the end of 2026.
The same page restates the post-Omnibus high-risk timeline as fact: following the political agreement on the AI Omnibus, the application of the high-risk rules was postponed to December 2027 for stand-alone AI systems and to August 2028 for AI embedded in products, so that providers and deployers have time to prepare before the legal deadline.
The guidelines are intended to help providers, deployers and market-surveillance authorities assess whether an AI system is high-risk, and are also presented in an “AI Act Single Information Platform” explorer with summaries and examples. The AI Act identifies two types of high-risk AI systems: those embedded in products regulated under Union product-safety law, and those in specific Annex III use cases that can significantly affect health, safety or fundamental rights.
Notable quotes
“The consultation was originally open for 6 weeks until 23 June. However, the Commission has received requests from several associations and other stakeholders to extend it further by 4 weeks. To give stakeholders additional time to submit their input, the deadline was extended to 23 July 2026.” — consultation page
“The final guidelines will be adopted by the end of 2026.” — consultation page
What’s speculative vs. asserted
- Asserted (primary, this scan): consultation opening 19 May 2026; deadline extended from 23 June to 23 July 2026; final guidelines adopted end-2026; Omnibus postponement of high-risk rules to Dec 2027 (stand-alone) / Aug 2028 (product-embedded).
- Not asserted: any change to the substance of the draft classification examples; any change to the 2 August 2026 GPAI / Article 50 track (out of scope of this page).
Topics this feeds
- EU AI Act — resolves the standing 23 June vs 23 July consultation-deadline tension with primary evidence.
- EU AI Office (European Commission) — the facilitating body; tracked-change updated.
Open questions raised
- Will the final high-risk guidelines change the worked examples materially from the 19 May 2026 draft, and will any examples touch financial-services use cases (credit scoring, fraud, AML)?
- Given end-2026 adoption is after the 2 August 2026 core-obligations date (but before the Dec 2027 high-risk application date), how will firms treat the draft guidelines as an interim classification reference?