EU AI Act guidance puts high-risk systems in focus

Summer 2026 has brought two major steps in the implementation of the EU AI Act, sharpening expectations for organisations that deploy AI. The European Commission’s draft guidelines, published 19 May 2026, clarify when AI systems are considered high-risk, including systems embedded in regulated products such as medical devices and machinery, and those used in sensitive areas such as employment, education, biometrics or access to essential services.

The guidance stresses that classification should happen at the level of each use case, based on both intended purpose and actual use. Human sign-off alone is not enough to avoid high-risk classification, and the guidelines, while not legally binding, are expected to inform enforcement by regulators and market surveillance authorities.

Ireland is also moving toward domestic enforcement. On 17 June 2026, the Irish Government approved publication of the Regulation of Artificial Intelligence Bill 2026, which establishes Oifig IS na hÉireann as the central coordinating authority and empowers Market Surveillance Authorities to supervise compliance, investigate breaches and impose sanctions. Organisations are being advised to inventory AI use cases, assess risk, assign governance ownership and align documentation with anticipated supervisory requirements.