The EU AI Act Is Now in Force and Irish Companies Have Until August to Get Compliant

Why It Matters

Non‑compliance will trigger substantial fines and erode procurement eligibility and partnership trust across the EU market.

Key Takeaways

• High‑risk AI includes recruitment, credit scoring, and education assessments
• Penalties reach up to €35 M ($38 M) or 7% turnover
• Compliance requires conformity assessment, CE marking, and EU database registration
• Irish fintech overseen by Central Bank; HR tools by Workplace Relations Commission
• Start an AI inventory now to meet August 2 deadline

Pulse Analysis

The EU AI Act, the first continent‑wide framework governing artificial intelligence, entered its enforcement phase on August 2 2026. By categorising systems into prohibited, high‑risk, limited‑risk and minimal‑risk tiers, the legislation targets applications that directly affect individuals’ rights—such as hiring algorithms, credit‑scoring models, and educational assessment tools. Ireland’s concentration of global tech headquarters and a burgeoning AI startup ecosystem makes the country a focal point for the Act’s impact, as a disproportionate share of EU‑based AI activity will be evaluated against these new standards.

For high‑risk AI, the compliance burden is substantial. Providers must complete a conformity assessment, issue an EU declaration of conformity, affix the CE mark, and register the system in the EU’s central database. Ongoing duties include documented risk‑management, robust data‑governance, automatic logging, and human‑oversight mechanisms, all subject to inspection by sector‑specific regulators—FinTech firms answer to the Central Bank, while HR platforms report to the Workplace Relations Commission. Penalties can climb to €35 million ($38 million) or 7% of worldwide turnover for prohibited practices, underscoring the Act’s enforcement teeth. The regulated online‑gaming market offers a useful parallel, having long required transparency, audit trails and consumer‑focused safeguards that now serve as a benchmark for AI governance.

Irish companies should treat the August deadline as a catalyst for immediate action. The first step is an AI inventory: a detailed catalogue of every AI system, its function, and a preliminary risk classification. High‑risk candidates then undergo a deeper review against Articles 8‑15, ensuring documented governance, human oversight and conformity evidence are in place. Guidance from KPMG, EY and the Irish government converges on the view that compliance will become a prerequisite for procurement and partnership eligibility, not merely a legal checkbox. Early adoption of these practices will differentiate trustworthy vendors in a market where regulatory certainty is rapidly becoming a competitive advantage.