Key Takeaways
- Fewer than half of Irish Government departments and agencies had rolled out artificial intelligence training by early 2025.
- The approach to AI varies widely, with some State bodies embracing the technology while others remain cautious or avoid its use.
- Just over half of these organisations indicated they have, or are developing, specific AI policies.
- New government guidelines aim to support the ethical adoption of AI across public services.
- There’s a call for greater AI use to improve efficiency and free up public servants for policy work.
Irish government departments and agencies are showing a mixed response to the rapid advancements in artificial intelligence. As of early 2025, fewer than half had provided AI training to their staff, indicating a varied pace in adapting to this new technology.
A survey of 18 Government departments and 40 other official bodies at the end of January revealed this patchwork approach. Some are actively exploring AI, while others are holding back, at least for now, according to information obtained by The Irish Times.
In an effort to guide this transition, the Department of Public Expenditure recently published updated guidelines on the ethical use of AI by civil servants. Minister for Public Expenditure Jack Chambers highlighted AI’s “immense possibilities” for public services, stressing the need to “embrace that change and adapt quickly.”
Despite new resources, including a framework for evaluating AI use and free courses, many agencies have so far relied on broader guidance documents. The survey showed that specific, tailored AI policies exist or are planned in about half of government departments and just over half of State agencies.
Malcolm Byrne, a Fianna Fáil TD who is set to chair Ireland’s new Oireachtas committee on AI, expressed hope that the new guidelines would encourage a shift. He noted some positive examples of AI use but was concerned by an “abundance of caution” in many responses.
Mr. Byrne believes public services should use AI for efficiency gains, allowing staff to focus more on policy issues rather than mundane tasks. His comments came after reviewing the survey responses, which paint a picture of varied official views on AI.
For instance, fourteen Government departments have ruled out using AI for drafting legislation. The Department of Finance, following advice from the National Cyber Security Centre (NCSC), stated that generative AI (GenAI) – which creates original content from prompts – should not be used for official correspondence or policy design.
Similarly, the Agri-Food Regulator and the National Transport Authority (NTA) prohibit the use of GenAI, though the NTA acknowledges its potential benefits. Staff at the Garda National Data Protection Office have received training on AI risk assessment for data protection.
However, some bodies are embracing the technology. The Loughs Agency is developing AI to help validate its fish counters. The Health Service Executive (HSE) has established a working group for AI implementation in public health.
Teagasc, the agriculture and food development authority, reported using large language AI models, while also acknowledging the need to manage associated risks. This contrasts with many other agencies that currently restrict or avoid such tools, highlighting the diverse strategies across the Irish public sector as it navigates the implications of AI and regulations like the EU AI Act, which came into force in August 2024.
