Regulatory explainer · AI Regulations and Compliance

EU AI Act: enterprise readiness guide

A working checklist covering classification, roles and governance implications, based on the timetable as amended in 2026.

On this page
  1. Regulatory overview
  2. The risk-based approach
  3. Organisational responsibilities
  4. AI system classification
  5. Governance implications
  6. Practical preparation steps

Educational content, not legal advice. This explains publicly available material as of the publication date. Regulatory requirements change, so confirm how they apply to your organisation with qualified counsel.

Regulatory overview

The AI Act (Regulation (EU) 2024/1689) entered into force on 1 August 2024, and Regulation (EU) 2026/1744 amended it with effect from 27 July 2026. High-risk rules for Annex III areas now apply from 2 December 2027, and those for products under Annex I from 2 August 2028. Prohibitions, general-purpose AI rules and Article 50 transparency duties are already in application. See our EU AI Act explainer for the full timeline.

The risk-based approach

The Act sets out four levels: unacceptable risk (banned), high risk (strict obligations), transparency risk (disclosure duties) and minimal or no risk (no rules under the Act). General-purpose AI models have separate obligations.

Organisational responsibilities

  • We know where we act as provider, deployer, importer or distributor.
  • We have checked whether customising vendor AI could make us a provider under Article 25.
  • Someone is accountable for AI Act readiness, with support from legal, security and data teams.

AI system classification

  • Every AI system, including vendor features, is in our inventory.
  • We have recorded whether each system or its output reaches the EU.
  • We have checked each use against Article 5 prohibitions, including the 2026 addition.
  • We have checked each use against the Annex III areas and Annex I products.
  • We have documented our reasoning, including any reliance on Article 6(3).

Governance implications

  • Our AI governance policy reflects the Act’s roles and risk levels.
  • High-risk deployers have assigned competent human oversight and log retention.
  • Supplier contracts secure the documentation and cooperation we need.
  • AI reviews are aligned with GDPR data protection impact assessments.
  • Staff using AI receive suitable AI literacy support.

Practical preparation steps

  1. Now: confirm no prohibited practices, and address Article 50 transparency duties.
  2. Next: complete the inventory and classification, and agree ownership.
  3. Then: build risk management, documentation and oversight for high-risk systems ahead of the December 2027 and August 2028 dates.
  4. Ongoing: track harmonised standards, Commission guidance and national enforcement.

Sources

EU AI ActAI regulatory complianceAI compliance

Share: LinkedInXEmail

Talk to Cognexaa

Knowledge creates confidence. Technology makes it possible.

Talk to our team about how Cognexaa’s platforms can help your organisation approach AI governance, cybersecurity and digital trust with greater structure.

Or email us directly at sales@cognexaa.com