Ireland Is About to Set the EU’s AI Agenda. It Also Hosts the Industry.
On July 1, Ireland will take the rotating presidency of the Council of the European Union. For six months, it will chair negotiations, set agendas, and broker compromises on legislation affecting 450 million people. One of the biggest items on the table is how and when the EU enforces the rules it already passed to govern artificial intelligence.
Ireland is also where Anthropic just expanded its European headquarters sixfold, adding 200 employees in Dublin. OpenAI has been growing its Irish operations. Workday opened a $204 million AI research center there last year. Taoiseach (Prime Minister) Micheál Martin has made attracting American AI investment a centerpiece of the country’s economic strategy, and Ireland’s foreign direct investment arm, IDA Ireland, is positioning the country as the industry-friendly gateway to Europe. In October, Ireland will host an International AI and Digital Summit in Dublin, branded around “Enabling AI to Power European Growth.”
None of this is accidental. Ireland has been running this playbook since the 1990s, when it began using tax policy and talent investment to attract American tech firms. The AI push is the latest iteration, not a new strategy. The IMF recently ranked Ireland first globally for skill readiness. I did my master’s research on algorithmic bias at University College Dublin, and what I learned living and researching there was that Ireland builds these positions deliberately, over decades, through coordinated policy.
The tension is in the timing. The EU Council agreed in March to delay enforcement of its AI Act’s high-risk system rules from August 2026 to December 2027 for standalone systems and August 2028 for systems embedded in products. That gives companies building AI for hiring, credit scoring, law enforcement, and healthcare up to two extra years before they face compliance obligations. And because the AI Act is not retroactive, systems placed on the market before those deadlines may never have to comply at all. Laura Caroli, a former co-negotiator of the AI Act, warned that an AI hiring tool launched before December 2027 “may remain outside the AI Act indefinitely.”
The delay happened after a significant lobbying push. A Corporate Europe Observatory and LobbyControl analysis found that 69% of European Commission meetings on the Digital Omnibus simplification package in 2025 were with business groups. Only 16% were with NGOs. Digital lobbying spending in Brussels reached €151 million, up 33% in two years. In one AI policy consultation, researchers found only 11 or 12 participants, all from industry except for a single civil society organization. That climate matters because the Council presidency does not determine outcomes alone, but it does shape which compromises become politically possible during its six-month window.
IDA Ireland has described the country’s Council presidency agenda as advancing “simplification” of EU digital rules. Ivan Houlihan, IDA’s head of the western U.S., told the Washington Post that Ireland would focus on making sure regulations have “no unintended consequence” for AI companies. That is a reasonable position for a country whose economy depends on attracting exactly these firms. But the country setting the EU’s regulatory agenda for six months also has a clear economic incentive to keep enforcement flexible.
What makes Ireland worth watching is that it has also shown real governance ambition on AI. The country’s own Joint Committee on Artificial Intelligence published a report in December 2025 stating that the EU AI Act should be treated as “a minimum baseline for national AI regulation, not a maximum standard.” Ireland has designated 15 specialized enforcement authorities for AI Act compliance and is building a National AI Office by August 2026 with a built-in regulatory sandbox for testing new systems. I contributed to the CAIDP AI and Democratic Values Index 2026 as a research group member, evaluating how countries align AI policies with democratic governance principles. The kind of institutional groundwork Ireland is laying is exactly what that framework is designed to measure.
The question for those six months is which version of Ireland leads the Council. The one that told its own parliament the AI Act is a floor, not a ceiling. Or the one that told American investors it wants to make sure regulation creates no friction for the companies it is working to attract. Both positions are defensible on their own. Holding them at the same time, while chairing the body that decides how fast enforcement moves, is the real governance test.