456. Deputy Shónagh Ní Raghallaigh asked the Minister for Enterprise, Tourism and Employment whether the Government intends to introduce a regulatory framework governing artificial intelligence tools that market themselves as mental health, counselling, or psychotherapy services; and if he will make a statement on the matter. [64482/26]

I thank the Deputy for the question.

As the Deputy will be aware, Government has introduced a national regulatory framework through the Regulation of Artificial Intelligence Act 2026 (No. 31 of 2026), which provides for the supervision and enforcement in Ireland of Regulation (EU) 2024/1689, the EU AI Act. The framework includes the recently established AI Office of Ireland, which promotes the safe, responsible and trusted use of artificial intelligence across Ireland. It works alongside the relevant sectoral competent authorities responsible for monitoring and enforcement under Ireland’s distributed regulatory model.

The EU AI (AI) Act, which came into force in 2024 with its provisions applying in a phased manner over the next two years, establishes a horizontal, risk-based framework for the regulation of AI developed or used in the EU. The Act aims to promote adoption and support innovation whilst also protecting health, safety and fundamental rights, including democracy and the rule of law. The Act categorises AI systems by the level of risk they pose and imposes corresponding obligations.

In Article 5, the AI Act prohibits AI practices that pose an unacceptable risk, including certain manipulative, exploitative, social-scoring and biometric uses.

Chapter III of the Act classifies certain systems as high-risk where they may affect health, safety or fundamental rights, and subjects them to strict requirements on risk management, data, documentation, human oversight, accuracy and cybersecurity.

Chapter V of the Act recognises the special role general-purpose AI (GPAI) models play in the AI ecosystem and sets out stringent requirements for providers of GPAI models to comply with throughout the lifecycle of the model, to be supervised and enforced centrally by the European Commission.

Article 50 requires transparency where people interact with AI or encounter specified AI-generated or manipulated content, including clear labelling of deepfakes. The GPAI systems used for therapy or emotional support must comply with the transparency obligations and make it clear to users that they are interacting with an AI system.

I am cognisant of the fact that AI tools can significantly enhance the health and wellbeing of our citizens. But it can also cause harm, especially if it not designed or approved for this purpose. These risks are worrying for vulnerable users, including children. I am aware that the use of AI systems, particularly general-purpose AI systems for therapy or emotional support is a growing phenomenon. These AI systems, and the underlying general-purpose AI models, must comply with their obligations under the AI Act.

The Regulation of Artificial Intelligence Act 2026 was commenced in July and puts in place the national institutional and enforcement architecture required to apply the EU AI Act's provisions and provides a comprehensive legal framework for ensuring that the AI sector in Ireland develops in a responsible and ethical manner. Central to this national framework is the principle that the AI systems which pose risk to our citizens health and wellbeing must demonstrate their safety and compliance before they enter the market.

I want to assure the Deputy that this Government is committed to AI development that is safe, human-centric and trustworthy. The combination of EU and national legislation, including the EU AI Act and the Regulation of Artificial Intelligence Act 2026, along with the establishment of the AI Office of Ireland and the designation of sectoral regulators, provides a comprehensive and proportionate framework which will allow us to benefit from AI while safeguarding our citizens from its harmful effects.