AI Training for HR Teams in Ireland: What Your People Team Needs to Know

HR teams sit at an awkward intersection with AI. They are asked to adopt tools for recruitment, onboarding, performance and employee queries, and they hold the most sensitive personal data in the organisation. They are also working in the part of the EU AI Act that regulators single out as high-risk. This guide sets out what AI training for HR teams in Ireland should cover, based on the EU AI Act text as shown on the European Commission's AI Act Service Desk, Data Protection Commission (DPC) guidance, gov.ie and CIPD publications. It applies equally to people teams in Dublin and elsewhere, because the rules are national and EU-wide.

Two companion pieces go deeper on specific points: our guide to AI in recruitment and HR under the EU AI Act covers hiring tools, and our explainer on the AI literacy duty in Article 4 covers the general training obligation. This article focuses on the content of a training programme for people teams. It is general information, not legal advice.

Why HR is a high-risk area

Annex III of the AI Act, as shown on the Commission's AI Act Service Desk, lists employment, workers' management and access to self-employment as a high-risk area. The uses it names include placing targeted job adverts, screening and analysing applications, evaluating candidates, making decisions that affect working conditions, promotion or dismissal, allocating tasks based on behaviour, personal traits or characteristics, and monitoring and evaluating the performance and conduct of workers. The page shows no date of its own.

The gov.ie page "The EU AI Act and my organisation" (published and last updated 27 May 2026) says high-risk systems, with hiring given as an example, carry strict obligations for documentation, transparency and human oversight. CIPD's Ireland briefing on the Act (30 June 2025) likewise says employment is a high-risk area covering tools such as CV-sorting software and systems used to manage workers. Whatever the final timetable, HR is plainly in scope.

What the law says about training

Two provisions matter. First, Article 4 (AI literacy). The Service Desk page, based on the consolidated text as at 27 July 2026, shows all three paragraphs of Article 4 marked as amended by the Digital Omnibus on AI. Its paraphrase says providers and deployers must take steps to build AI literacy among staff and others operating or using AI on their behalf, taking account of technical knowledge, experience, education and training, the context of use and the people affected, and that the obligation does not require any individual to reach a particular level. The page's non-binding summary says providers and deployers "should take measures to support the development of AI literacy" of their staff, and adds that the summaries are not legally binding. The Commission FAQ describes the Omnibus as a proposal, says AI literacy rules already apply and says training obligations for high-risk deployers would remain. Because the pages differ, check the current text of Article 4 and treat training as expected either way.

Second, Article 26 on deployers of high-risk systems. According to the Service Desk page, oversight must be assigned to people who have the necessary competence, training and authority, and who have the support they need. Deployers must also, before using a high-risk system at work, inform workers' representatives and the affected workers, and inform individuals subject to decisions made or assisted by an Annex III system that the system is being used. A person in HR is often the one who will do the informing. Our guide to the DPC's AI and large language model guidance covers the data protection side in more detail.

What an HR training programme should cover

What follows is our suggestion for a curriculum, mapped to the sources above. It is not a prescribed syllabus.

1. What counts as AI, and where it hides

Staff should be able to recognise AI features in tools they already use: applicant tracking systems, video interview software, chatbots, scheduling and rostering tools, sentiment or engagement analytics, and generative assistants. The gov.ie page recommends mapping where AI is used, including vendor tools, and classifying each system by risk. HR should know where that inventory lives and how to add to it.

2. Risk categories and banned practices

People teams should know the Act's tiers as described by gov.ie: prohibited, high-risk, limited-risk such as chatbots with transparency duties, and minimal-risk. One prohibition is directly relevant to HR. The Service Desk page for Article 5 says the Act prohibits AI systems placed on the market, put into service or used for the specific purpose of inferring a person's emotions in the workplace or education institutions, except where the system is intended for medical or safety reasons. gov.ie lists emotion recognition in workplaces among prohibited practices, and CIPD's briefing mentions it too. An HR manager shown a tool that claims to read candidates' or employees' emotions needs to know to stop and escalate.

3. Human oversight and automation bias

The Service Desk page for Article 14 describes oversight that lets people understand the system's capabilities and limits, remain aware of automation bias, interpret output correctly, and decide not to use, override or reverse it. Training should include realistic exercises: reviewing a ranked shortlist, challenging a recommended performance rating, and documenting a decision to depart from the system's output. CIPD's March 2026 guide for people professionals says the people team should check whether human oversight is meaningful, and look for bias, impact assessments before deployment, clear monitoring roles, transparency and escalation routes.

4. Data protection with AI

The DPC's page "AI, Large Language Models and Data Protection" (18 July 2024) lists points for organisations using AI products: consider a formal risk assessment, check that the provider's documentation is understandable and accessible, watch for unwanted processing that could breach lawfulness, fairness, transparency and purpose limitation, avoid using more personal data than needed, be able to handle access and deletion requests, watch for inaccurate or biased outputs, and note that purely automated decisions can cause harm, so critical human review matters. HR teams should be able to apply these to a real tool. A practical rule, labelled as ours: do not paste candidate or employee personal data into a public generative AI tool unless it has been approved.

5. Employee rights and explanations

Article 86, as summarised on the Service Desk, gives a person affected by a decision based on an Annex III system the right to clear and meaningful explanations of the AI system's role in the decision, where the decision has legal effects or similarly significant effects and they consider it adversely affects their health, safety or fundamental rights. HR staff should know how to respond to such a request and who to involve.

6. Governance and speaking to the business

CIPD's guide says wherever AI has a human impact, the people team needs to be as involved as IT, legal and compliance colleagues, and suggests an AI policy or charter and involving employee representatives. Training should prepare HR to take part in governance discussions, not simply receive decisions.

Who needs which level of training?

Article 4 refers to technical knowledge, experience, education, context of use and affected people, so a single course for everyone is unlikely to fit. A practical approach: give everyone in the team a short baseline on the points above, give recruiters and HR business partners using high-risk tools a deeper module on oversight and notices, and give HR leaders a governance module. Keep a record of who completed what, and refresh it when tools change.

On enforcement, the AI Office of Ireland's list of national competent authorities includes the Workplace Relations Commission, described as supervising the deployment and usage of AI systems in the Irish labour market, and the Data Protection Commission, described as a market surveillance authority for certain prohibited practices and high-risk uses, with short descriptions only.

Learnsignal's offices are off Baggot Street in Dublin, and we offer in-person training from there. For current details, see our CPD pages.

Frequently asked questions

Is AI training mandatory for HR teams in Ireland?

Article 4 deals with AI literacy for providers and deployers, and gov.ie says staff who develop, deploy or manage AI should receive training on the technology and the legal obligations. The wording of Article 4 is shown as amended on the Commission's page, so check the current text.

Does the AI Act ban AI in recruitment?

No. Annex III treats recruitment tools as high-risk, which means obligations rather than a ban. The Act does prohibit systems designed to infer emotions in the workplace, other than for medical or safety reasons.

Do we need a data protection impact assessment for an AI HR tool?

In its section for AI developers and providers, the DPC says building an AI product may require a DPIA and that doing one anyway may be best practice; for organisations using AI products it says a formal risk assessment should be considered. The Article 26 page also points to using the provider's information for a GDPR DPIA where applicable.

When do the high-risk rules apply?

The Commission's consolidated text and timeline pages show 2 December 2027 for Annex III systems. The position has been subject to the Digital Omnibus on AI, so confirm the current date before planning.