Today’s Focus: AI Governance
Today’s Top Story
EU Digital Omnibus Trilogue Targets 28 April Deal
The Cypriot Presidency is racing toward political agreement at the second AI Act trilogue on 28 April 2026 — just 13 days away. The Council adopted its mandate on 13 March and Parliament confirmed its position on 26 March, both backing amendments that would push stand-alone high-risk AI obligations to 2 December 2027 (and 2 August 2028 for AI embedded in regulated products), while locking in provider watermarking for synthetic audio, image, video, and text by 2 November 2026. If the trilogue lands, your high-risk AI compliance roadmap and your generative-content labeling timeline both move — in opposite directions.
Sources: https://digital-strategy.ec.europa.eu/en/library/digital-omnibus-ai-regulation-proposal · https://www.europarl.europa.eu/legislative-train/package-digital-package/file-digital-omnibus-on-ai
Also Today
Colorado AI Act enforcement starts in 76 days
SB 25B-004 locked the new effective date at 30 June 2026 without softening substantive duties. Deployers of high-risk AI must have risk-management policies, initial and annual impact assessments, pre-decision and adverse-decision consumer notices, and website disclosures live before the bell.
Action: Finalize your impact-assessment template and notice flows this month. Source: https://www.hunton.com/privacy-and-cybersecurity-law-blog/enforcement-of-colorado-ai-act-delayed-until-june-2026
China’s anthropomorphic-AI measures take effect 15 July 2026 (91 days)
Five agencies (CAC, NDRC, MIIT, MPS, SAMR) released the Interim Measures for AI Anthropomorphic Interaction Services on 10 April 2026. Virtual companions and emotional chatbots must add anti-over-dependence safeguards, AI-disclosure notices, and must not be offered to minors.
Action: Confirm in-scope features and stand up minor-detection blocking now. Source: https://www.chinalawtranslate.com/en/chatbot-measures-draft/
NY RAISE Act chapter amendments signed
Governor Hochul finalized the frontier-model transparency law on 27 March 2026; effective 1 January 2027. In scope: developers training models above 10²⁶ FLOPs and >$100M compute cost with annual revenue over $500M. Required: published safety protocols, third-party audits, 72-hour incident reporting to the NY AG and DHSES, with civil penalties up to $1M (first) and $3M (subsequent).
Action: Map your training-compute and revenue against scope today. Source: https://www.governor.ny.gov/news/governor-hochul-signs-nation-leading-legislation-require-ai-frameworks-ai-frontier-models
FTC Policy Statement on AI and Section 5 (11 March 2026)
The Commission interprets “unfair or deceptive acts or practices” as reaching algorithmic discrimination, deceptive AI-generated content, and AI-related privacy violations across the lifecycle — and signals it views some state laws compelling alteration of “truthful AI outputs” as preemptable.
Action: Refresh AI marketing and product-claims review against the new examples. Source: https://ourtake.bakerbotts.com/post/102mirs/march-2026-federal-deadlines-that-will-reshape-the-ai-regulatory-landscape
Deadline Alert
EU Digital Omnibus trilogue — 28 April 2026 (13 days). Watch for political agreement that resets your AI Act high-risk timelines. If you’re holding back compliance investment expecting delay, validate against the actual final text the morning after.
One Thing to Do Today
Put 28 April 2026 on your AI compliance calendar with a same-day workstream review trigger. If the trilogue lands, you’ll need to immediately rebaseline your EU AI Act high-risk roadmap to a 2 December 2027 target and your watermarking program to 2 November 2026 — both materially different from the status quo.
Tomorrow’s Focus
Thursday is CyberEyeQ Weekly — your cross-domain synthesis of the week’s most important regulatory moves.
CyberEyeQ — Actionable Regulatory Intelligence. Questions, tips, or feedback? Reply to this email or write to [email protected].