5th March 2025
On Tuesday, 4 March 2025, the government approved a recommendation from Minister for Enterprise, Tourism and Employment, Peter Burke, that Ireland adopt a distributed model of implementation of the EU Artificial Intelligence (AI) Act. This approach will build on the deep knowledge and expertise of the established sectoral regulators. The government approved the designation of an initial list of eight public bodies as competent authorities, responsible for implementing and enforcing the Act within their respective sectors. These authorities are:
- Central Bank of Ireland
- Commission for Communications Regulation
- Commission for Railway Regulation
- Competition and Consumer Protection Commission
- Data Protection Commission
- Health and Safety Authority
- Health Products Regulatory Authority
- Marine Survey Office of the Department of Transport
Additional authorities, and a lead regulator who will coordinate enforcement of the Act and provide a number of centralised functions, will be designated by a future government decision to ensure comprehensive implementation of the Act.
Minister for Enterprise, Tourism and Employment, Peter Burke said,
“AI presents Ireland with a strategic opportunity; it holds the prospect of major benefits for our economy and for our society. For business it can boost productivity, spur innovation and deliver better customer services; for the public it can provide enhanced public services; and for society, accelerated advances in science and medicine. It is a priority for me to ensure that we capture these benefits.
“However, to capture these benefits, we must build trust in AI systems. For this reason, the landmark EU AI Act, the first in the world comprehensive regulation establishing guardrails for the safe and ethical use of AI, is a strategically important regulation for Ireland, as well as the EU. I am committed to an efficient and well-resourced implementation of the Act in Ireland, in a manner that provides the necessary safeguards, while spurring innovation for the benefit of our economy and our society.”
Minister of State for Trade Promotion, Artificial Intelligence and Digital Transformation, Niamh Smyth said,
“The decision by government to use the existing national framework of well-established sectoral authorities for enforcement of the EU AI Act will make compliance with the AI Act easier for businesses. It is also an important step towards the commitment in the Programme for Government to make Ireland an EU centre of expertise for digital and data regulation for companies operating across the EU Digital Single Market. Providing an efficient, comprehensive, fair and transparent implementation of the Act in Ireland will enhance Ireland’s reputation for quality regulation and its competitiveness for attracting further investment in this burgeoning technology.”
ENDS
For Editors
The EU AI Act establishes a harmonised regulatory framework for AI systems developed or deployed in the EU. It is designed to provide a high level of protection to people’s health, safety, and fundamental rights and to simultaneously promote the adoption of human-centric, trustworthy AI. The Act entered into force in August 2024 and its provisions apply, in a phased manner, over the period to August 2027.
The Act is a horizontal instrument that applies to all sectors of the economy, both public and private. However, there are exemptions for applications of AI relating to national defence; national security; scientific R&D; R&D for AI systems, models; open-sourced models; and personal use.
The Act is risk-based so that its provisions are targeted and proportionate – it is not a blanket instrument applying to all AI systems. Most AI systems are not subject to any regulatory requirements under the Act as they are low risk. In addition, the Act gives special consideration to the needs of SMEs and startups. This will ensure that the EU remains competitive for AI investment and innovation. The key elements of the Act are as follows:
- Eight AI practices are prohibited from February 2025 due to the unacceptable risk they pose:
- Subliminal techniques likely to cause that person, or another, significant harm
- Exploiting vulnerabilities due to age, disability or social or economic situation
- Social scoring leading to disproportionate detrimental or unfavourable treatment
- Profiling individuals for prediction of criminal activity
- Untargeted scraping of facial images
- Inferring emotions in workplaces or education institutions
- Biometric categorisation of race, religion, sexual orientation…
- Real-time remote biometric identification for law enforcement…
- Stringent conditions must be satisfied by high-risk AI systems, by their providers, and by their deployers, in order for such systems to be placed on the market or put into use. The Act identifies two classes of high-risk systems:
- AI systems that are part of the safety components of twelve specific product categories e.g. toys, machinery (applies from August 2027).
- AI systems in eight specific uses, for example, employment, education, in relation to essential public and private services such as financial, healthcare (applies from August 2026).
- Transparency conditions are placed on providers and deployers of four categories of AI systems that give rise to lower-order risks, such as chatbots (applies from August 2026).
- Providers of General Purpose AI (GPAI) models (foundation models) are subject to obligations to mitigate the substantial risks, including systemic risks, they pose due to their power and generality. These obligations will be enforced by the European Commission, but with the cooperation of member states (applies from August 2025).
- The penalties for infringements of the Act are substantial: fines of up to €35 million or 7% global turnover
Back to Department News