AI Literacy for Healthcare Staff in Ireland: What Article 4 Means for Health Employers

If your hospital, nursing home, disability service, home support provider or GP or dental practice uses AI tools, Article 4 of the EU AI Act is relevant to you. It asks providers and deployers of AI systems to take measures to support the AI literacy of their staff. This guide explains what the official sources say about AI literacy for healthcare staff in Ireland, what they leave open, and where HIQA's new guidance fits. It is a workforce and compliance guide, not clinical advice, and not legal advice.

For the general employer duty, start with our guide to Article 4 and the AI literacy training duty for Irish employers. This post applies that duty to health and social care settings.

What Article 4 says now

The wording changed in 2026. EUR-Lex shows Regulation (EU) 2026/1744, the Digital Omnibus on AI, as adopted on 8 July 2026 and published in the Official Journal on 24 July 2026. It replaces Article 4. Under the new text, providers and deployers must "take measures to support the development of AI literacy of their staff". The text adds that the obligation "does not require providers or deployers to guarantee any specific level of AI literacy of any individual."

The European Commission's AI literacy Questions and Answers page (last updated 27 July 2026) explains how to read this. It says the obligation covers staff and other persons dealing with AI systems on the organisation's behalf, which can include contractors and service providers. Measures should take account of differences in staff knowledge and the context in which the systems are used.

The same page makes several points that matter in a regulated sector:

• There is no one-size-fits-all approach and no strict or mandatory training requirement.
• Relying only on a system's instructions for use is described as usually insufficient.
• Organisations are not required to measure employees' AI knowledge, and no certificate is needed. They can keep internal records of training or other initiatives.
• No particular governance structure, such as an AI officer or board, is required.
• Deployers of high-risk systems must still ensure staff are trained for human oversight.

Healthcare and sector-specific requirements

Health employers may wonder whether stricter AI literacy rules apply to them. The Commission page says the AI Office does not impose requirements for specific sectors, and that the sector and purpose of use should still inform an AI literacy initiative. Our reading: the duty is the same in principle for a Dublin hospital group and a small practice; what differs is the context, the systems in use and the risk.

The page says that if an organisation's AI systems are high-risk, additional measures might be relevant. Recital 91 of the original AI Act text on EUR-Lex says deployers should ensure that the people assigned to human oversight have "the necessary competence, in particular an adequate level of AI literacy, training and authority". The recital is explanatory and the Article 4 wording has since been revised, so read the two together and check the current position.

Who enforces it, and what the sources say about dates

According to the Commission page, supervision and enforcement sit with national market surveillance authorities, not the AI Office, and penalties are based on national law. It adds that enforcement is proportionate, that factors include the nature and gravity of the infringement and whether it was intentional or negligent, and that a lack of appropriate training may make enforcement more likely after an incident. It also says the AI Act does not create criminal offences or a right to compensation.

The Commission page says Article 4 applied from 2 February 2025 and gives 2 or 3 August 2026 as the date supervision and enforcement begin, in different places, so check the official text. The Commission's AI Act regulatory framework page (last updated 3 August 2026) lists general application from 2 August 2026, with high-risk use-case rules for Annex III from 2 December 2027 and high-risk rules for regulated products from 2 August 2028. The Omnibus changes are recent; check the current position before planning.

Ireland uses a distributed model of national competent authorities, according to the AI Office of Ireland's page (no date shown). That page names the Health Products Regulatory Authority as monitoring the safety and performance of medical devices and does not list HIQA among the designated authorities. We therefore cannot tell you from the sources we read which Irish body would supervise Article 4 in a health setting; check the AI Office of Ireland pages.

Where HIQA's guidance fits

HIQA published its first national guidance on the responsible and safe use of AI in health and social care services on 29 July 2026. According to HIQA's announcement, it is for all services providing health and social care in Ireland, was commissioned by the Department of Health and developed with the Department, the HSE and other partners, and is organised around four principles: accountability, a human rights-based approach, safety and wellbeing, and responsiveness. It covers practical topics such as governance, transparency and public engagement. Our summary of HIQA's national guidance on AI in health and social care goes further.

Two cautions from the announcement itself. First, it does not describe the guidance as mandatory; it presents it as promoting good practice and supporting HIQA's existing national standards. Second, the page does not explicitly call for AI literacy programmes. It says the guidance aims to build awareness and good practice among services and staff, and it says AI use must comply with legal obligations including the EU AI Act. So HIQA's guidance and Article 4 are separate things that point in the same general direction.

Turning the duty into a plan

The following steps are our suggestions, not requirements. They are one reasonable way to meet a duty that the sources describe as flexible.

• List what AI you use. Include tools embedded in software you already own, tools staff use on their own initiative and AI used by contractors. You cannot support literacy for tools you do not know about.
• Group staff by how they meet AI. A clinician, an administrator, a rostering manager and a board member need different things. Our guide to tailoring AI training by role sets out how to do this.
• Cover the basics for everyone. The Commission suggests a general understanding of AI, including what the organisation uses and the opportunities and risks, and encourages attention to legal and ethical aspects.
• Add depth where oversight is involved. Anyone assigned to oversee a higher-risk system needs to understand what it does, where it can be wrong and when to step in.
• Write down what you did. Records are optional under the Commission's page, but they help you explain your approach if asked, particularly after an incident.
• Review when tools change. A new vendor or a new feature can change the training needed.

Staff time is a practical constraint in rostered services. A practical approach is to build short sessions into existing induction, mandatory training days and team meetings rather than creating a separate programme.

What the sources do not tell us

We did not find an Irish or EU source setting out an AI literacy curriculum for health staff, a required number of hours, or a certification. The Commission page says explicitly that none is needed. We also did not read HSE digital health pages on AI use. If your organisation is part of the HSE or funded by it, check your own governance and policies as well.

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

Does Article 4 apply to a small health practice?

The Commission's AI literacy page says the obligation covers providers and deployers of AI systems, and describes support for small and medium-sized enterprises under Article 4(2). Our suggestion: keep what you do proportionate to the tools you use and your people's backgrounds.

Do healthcare staff need an AI certificate?

No. The Commission page says no certificate is needed and organisations are not required to measure employees' AI knowledge. Internal records of training or other initiatives are an option.

Is there a different rule for healthcare?

The Commission page says the AI Office does not impose requirements for specific sectors. Context and risk still shape what is appropriate, and deployers of high-risk systems must make sure staff are trained for human oversight.

Is HIQA's AI guidance mandatory?

HIQA's announcement does not describe it as mandatory. It presents it as guidance that promotes good practice and supports existing national standards.

Have the dates changed?

Some AI Act dates have changed following the Digital Omnibus on AI. Article 4 has applied since 2 February 2025 according to the Commission, but the wording was revised in July 2026. Check the current official position.

Literacy is easier to plan when you know your tools and your roles, so begin with a simple inventory and build from there.