The Central Bank of Ireland and Artificial Intelligence: What Regulated Firms Should Know

If your firm is authorised by the Central Bank of Ireland and uses artificial intelligence, what does the Central Bank actually expect? The short answer from the pages we read is that there is no separate AI rulebook, but there are clear supervisory expectations that build on existing rules. This guide summarises what the Central Bank has said publicly, with dates, and separates that from our own suggestions. Accountability is a recurring theme, so it helps to read it alongside our guide to the Individual Accountability Framework. It is general information, not legal advice.

What we found, and what we did not

We looked for Central Bank speeches, outlook reports and other publications on AI. We found several, and the most detailed is the Regulatory and Supervisory Outlook report of February 2026. We did not find a dedicated AI rulebook or a binding AI-specific guideline in what we opened, and the pages we read do not describe one. This guide does not cover the Central Bank's discussion paper on the Consumer Protection Code and digitalisation.

The February 2026 Regulatory and Supervisory Outlook

This is the Central Bank's main published statement on AI so far as we read. The Governor's foreword is dated 26 February 2026. Points it makes include:

• AI is treated as both an opportunity and a risk. The report lists risks such as bias, misinformation, data loss and "black box" opacity, and says generative AI can make fraud easier.
• The report says the widespread adoption of third-party AI tools changes the risk landscape and calls for stronger model governance.
• It states that many AI risks "are not new and are already covered by existing regulations and standards."
• It sets out four supervisory standards: strategic alignment (AI should be appropriate for the specific business challenge), accountability and explainability (clear accountabilities, human oversight and explainability), proportionate governance (risk management matched to the scale and sensitivity of the use) and compliance (processes to meet EU AI Act obligations, including transparency requirements).
• Planned supervisory work includes building understanding of AI use across the value chain, including agentic AI, reinforcing governance and model-management expectations, and supporting national implementation of the AI Act provisions within the Central Bank's remit.

The report also links AI to operational and cyber resilience and to dependence on third-party providers. Separately, it says the Central Bank will assess how firms have implemented the Individual Accountability Framework. This guide does not cover the report's appendices or its sector sections beyond banking.

Earlier statements: 2024 and 2025

The same themes appear earlier. In a speech to the OECD/FSB Roundtable on Artificial Intelligence in Finance on 22 May 2024, Colm Kincaid, then Director of Consumer Protection, said firms should understand what business challenge they are addressing with AI and understand the technology they use. He added that supervisors would look for clarity of accountability for decisions made on its use, that data or algorithmic bias may lead to discriminatory treatment of customers, and that the Central Bank would monitor whether AI excludes people from services. He described the Central Bank's approach as consultative and as reinforcing existing standards in the AI context. The speech does not contain binding rules.

In remarks at the Fintech Ireland Summit on 28 November 2024, Deputy Governor Sharon Donnery treated AI as one example of innovation that brings both benefits and risks, and said the basics of good governance, risk management and delivering for consumers remain true however novel the product. She said boards and management bodies are responsible for all activities undertaken by the firm. The speech sets no AI-specific rules and does not mention the EU AI Act.

The February 2025 Regulatory and Supervisory Outlook said firms were still exploring generative AI, and that model risk was growing where advanced AI and machine learning add to "black box" concerns. It asked that firms be clear that their use of AI is appropriate to the business challenge, assign clear accountability for AI systems and decisions based on them, and consider interpretability, explainability, fairness and recourse for consumers. It said the Central Bank was likely to be designated a market surveillance authority under the AI Act for certain uses.

The AI Act link

The 2026 outlook says the Irish government has designated the Central Bank as an AI Act regulator for financial services, and that legislation to implement the AI Act in Ireland was targeted for August 2026 (check whether it has since been enacted). It lists AI used for creditworthiness and credit scoring, and for life and health insurance risk assessment and pricing, as high-risk. It also notes that the European Commission proposed simplification measures in November 2025, including moving the start date for high-risk requirements to December 2027. These are changes in progress at EU level, so check the current position on an official Commission or gov.ie page before relying on any date. The Central Bank's 2025 Innovation Hub update, which is undated on its cover but refers to 2026 events, also describes the Central Bank as designated and says it surveyed the banking and payments sector in the first quarter of 2026.

For what deployers of AI systems must do in practice, see our separate guide to AI Act deployer obligations and human oversight. Our Dublin readers should note that the AI Act rules and the Central Bank's expectations are national or EU-wide, not specific to Dublin.

Operational resilience and DORA

The Innovation Hub update reminds firms of their obligations under the Digital Operational Resilience Act for AI-related ICT risk, including keeping a clear view of ICT risks and being able to respond to and recover from disruptions. The 2026 outlook also asks firms to consider contingency options because of their dependence on AI and cloud providers concentrated in a small number of locations. The pages we read do not set out the detail of these obligations in the AI context, so go to the official DORA texts for that.

Accountability and competence

The Central Bank repeatedly ties AI to governance and clear accountability. The Individual Accountability Framework guide linked above explains how senior executive accountability works, and the 2026 outlook lists assessment of how firms implement that framework among its governance priorities. Separately, our guide to the Minimum Competency Code and CPD covers competence requirements. The Central Bank pages we read do not say how the Minimum Competency Code applies to AI, so do not assume a link beyond what the Central Bank states.

Practical steps (our suggestions)

The following are our suggestions, not Central Bank requirements:

• Keep a register of AI tools and uses in the firm, including third-party and embedded tools.
• For each use, write down the business problem it solves and why AI is appropriate, mirroring the strategic alignment standard.
• Name an accountable owner for each use and record what human oversight applies.
• Test for bias and unfair outcomes where customers are affected, and keep the results.
• Assess third-party AI providers as part of your ICT and outsourcing risk work.
• Brief staff who use AI on what they may and may not do, and keep a record.
• Check the Central Bank's publications page regularly, as expectations are still developing.

An IOSCO note for capital markets firms

A Central Bank markets update dated 8 June 2026 notes that IOSCO published a supervisory toolkit for AI use in capital markets on 25 May 2026 and opened an industry survey. The page states no Central Bank requirement and mainly republishes IOSCO materials, so treat it as context for firms in that sector.

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 the Central Bank have a specific AI rulebook?

Not in anything we opened. Its 2026 outlook says many AI risks are already covered by existing regulation and sets out four supervisory standards. Check the Central Bank's site for anything newer.

Is the Central Bank an AI Act regulator?

The 2026 outlook says the Government has designated it as an AI Act regulator for financial services. The detail of the national law should be checked on gov.ie.

Which AI uses are high-risk in financial services?

The 2026 outlook lists creditworthiness and credit scoring and life and health insurance risk and pricing. It does not claim to list every case, so read the AI Act text itself.

Who is accountable for AI use in a regulated firm?

The Central Bank says it wants clear accountability for AI systems and decisions, and Deputy Governor Donnery said in 2024 that boards and management bodies are responsible for all activities of the firm.

This area is moving, so confirm current expectations directly with the Central Bank and take advice where AI affects regulated activities.