- with Senior Company Executives, HR and Inhouse Counsel
- with readers working within the Law Firm industries
The Joint Committee on Artificial Intelligence has published its First Interim Report, marking a significant milestone in Ireland's approach to artificial intelligence governance. Released in December 2025, the report distils the Committee's findings from its first seven public meetings and sets out 85 recommendations spanning regulation, public engagement, sectoral impacts and fundamental rights protection.
The Committee, chaired by Malcolm Byrne TD, was established to examine and make recommendations on Ireland's approach to the development, deployment, regulation and ethical considerations of artificial intelligence. Its mandate encompasses ensuring that Ireland's approach supports economic growth, innovation, public trust and societal benefit while safeguarding rights and mitigating risks. The Committee has 24 calendar months to complete its work and intends to publish further interim reports before delivering a final report.
The National AI Office and Regulatory Architecture
A central theme of the report concerns the establishment and functioning of Ireland's National AI Office, which will serve as the single point of contact under Article 70 of the EU AI Act. The Committee welcomes the Government's commitment, articulated by Minister of State Niamh Smyth, to have the Office operational by August 2026.
The report addresses concerns raised by witnesses, including the Irish Council for Civil Liberties, regarding the independence of the National AI Office. Given that the Office will be housed within the Department of Enterprise, Tourism and Employment, the Committee recommends that the Office possess the necessary levels of independence, technical expertise and resourcing to ensure no conflicts arise between the State's support for industry, the State's own deployment of AI systems, and the design, implementation and enforcement of regulations.
The Committee further recommends that the National AI Office establish well-resourced advisory panels for routine collaboration with young people, older people and people with disabilities, together with civil society groups capable of providing expertise on AI's harms and benefits. A national AI risk register should be developed within the Office to identify and monitor systemic risks across sectors.
Regarding the nine authorities designated to safeguard fundamental rights under the EU AI Act, the Committee calls for dedicated, ring-fenced and multi-annual resourcing to reflect their new responsibilities. The report emphasises that additional funding should ensure these bodies have the human, technical and financial resources required to discharge their functions effectively.
Position on the EU AI Act
The report takes a clear position on Ireland's relationship with the EU AI Act. Citing Dr Patricia Scanlon, Ireland's AI Ambassador and Chair of the AI Advisory Council, the Committee recommends that Ireland must not shy away from the EU AI Act or attempt to dilute it. The Act should be treated as a minimum baseline for national AI regulation rather than a maximum standard.
This recommendation responds to concerns that commercial interests may exert pressure to weaken regulatory requirements in favour of innovation. The report draws an analogy with the implementation of the General Data Protection Regulation (GDPR), noting that clear and effective regulation provided clarity for operators and benefits for individuals without deterring investment or harming society. As witnesses from Research Ireland observed, companies seeking to operate in Europe will conform to European regulatory standards rather than maintain separate products for different jurisdictions.
The Committee calls for implementation and proper resourcing of existing regulations and strategies, including the EU AI Act, the Digital Services Act, the Data Protection Acts and online safety codes. This should encompass additional training and resources for competent authorities designated under the AI Act.
Regulatory Sandboxes
The report endorses the establishment of a regulatory sandbox scheme by August 2026, enabling companies and organisations to develop AI products safely within a controlled environment. Witnesses emphasised that any sandbox would require adequate resourcing, with suggestions including the possibility of partnering with another EU Member State and providing priority access for small and medium-sized enterprises.
Recommender Systems and Online Safety
The Committee identifies the propagation of harmful and hateful content propagated by AI-powered recommender systems as a significant regulatory gap that requires attention. The report recommends that recommender systems be switched off by default and that social media companies be prohibited from activating recommender algorithms for accounts used by children.
Further recommendations include obligations on platform owners to prevent the use of AI-driven recommender systems for misinformation campaigns aimed at destabilising society, and the introduction of mandatory algorithmic impact assessments for high-risk AI systems in public services.
Copyright and Creative Industries
The report identifies copyright as a notable regulatory gap. While a dedicated module on the creative industries is planned for future work, the Committee recommends strengthening the EU Copyright Directive to ensure that content cannot be used to train AI models without the consent of its creators. This issue is expected to receive more detailed treatment in subsequent reports.
Accountability and Transparency in Public Sector AI
The Committee places significant emphasis on accountability, traceability, transparency and explainability in AI deployment across Ireland. Where AI systems are used for decision-making, the report states they must be transparent, auditable and capable of demonstrating non-discrimination on the nine grounds established in Irish equality law.
Specific recommendations for the public sector include the requirement that all public bodies and semi-state entities using AI in public services publish annual, evidence-based reports detailing the benefits, disadvantages and any nequalities identified. These reports should be publicly accessible to ensure transparency and accountability.
The Committee recommends establishing a publicly accessible central register for all algorithmic systems used by the Government and public bodies. This register should include details relating to the developer, deployer and vendor of AI systems, as well as rocurement costs, performance metrics and findings from pre-deployment assessments.
The Department of Public Expenditure and Reform's Guidelines for the Responsible Use of AI in the Public Service should, according to the report, be amended to include a reference to the Public Sector Human Rights and Equality Duty. This Duty should serve as the core framework guiding public bodies in AI adoption, ensuring systems are rights-compliant from the outset.
Procurement
The report identifies accountability and transparency in public
procurement as a key concern. Any AI-powered tools or systems under
consideration by Government Departments or State bodies must follow
a procurement process and undergo risk assessment and bias testing
prior to deployment. The Department of Public Expenditure,
Infrastructure, Public Service Reform and Digitalisation should
provide clear guidelines covering both the use and procurement of
AI systems and services.
Bias, Inclusion and Equality
The Committee devotes substantial attention to bias within AI systems and the imperative of equitable and inclusive AI adoption. The report references incidents where AI has produced outputs based on discriminatory biases or stereotypes, including the example cited by the DPO Network of Google's Gemma model, allegedly downplaying women's physical and mental health issues compared with men's when generating and summarising case notes.
The Committee recommends that unbiased training data, human oversight, thorough research and detailed testing should be mandatory to avoid biased outcomes. Companies or public sector bodies that deploy biased or discriminatory AI systems should be regarded as being in breach of equality and non-discrimination laws and should be actively pursued and prosecuted by state agencies while remaining liable to individual litigation.
Equality by design should constitute a core element throughout all phases of AI product development. The report calls for pathways and mechanisms ensuring that underprivileged communities and marginalised groups are involved in the design and assessment of AI-powered tools and systems.
Energy and Environmental Considerations
Almost every stakeholder who engaged with the Committee expressed concerns about the energy implications of AI. The report notes testimony from Research Ireland warning that extrapolating current growth patterns forward would quickly reach a point where AI would consume more electricity than the planet generates.
The Committee recommends that issues surrounding energy in Ireland be resolved as an urgent priority, with plentiful, sustainable, and renewable energy made available. AI use and development must account for the State's international obligations to reduce overall energy use under the EU Energy Efficiency Directive and obligations under the Climate Act.
An energy council comprising full-time paid employees should be established to address current energy challenges and plan for future needs. Solutions to make hardware, software and processes around AI more environmentally friendly must be found.
Children and Young People
The report emphasises that children and young people's rights and wishes must be central to AI discussions, design, policies, strategies, and legislation related to AI. Drawing on testimony from organisations including the National Youth Council of Ireland, the Children's Rights Alliance, CyberSafeKids and BeLonG To, the Committee heard evidence of both benefits and harms.
Benefits identified include support for learning, assistance for teachers, helping young people find a sense of community and learn about their identity, and facilitating referrals to credible, human-run support services.
Harms and concerns include the environmental cost of AI, overreliance on technology, misinformation, chatbots presenting as fake companions, deepfakes, bias, stereotyping, harm to minority groups, LGBT+ individuals being outed, and harm through body image issues, eating disorders and instructions on self-harm.
The Committee heard from the Irish Traveller Movement regarding the vulnerability of Traveller children to hate-based harms online, with automated discrimination present in some large language models, algorithms and pages or bots established solely to stereotype Travellers negatively. BeLonG To provided evidence that LGBT+ young people suffer from AI perpetuating discriminatory stereotypes and from AI-generated content targeting minority groups online.
The report recommends that operators must be legally required to ensure that any product, platform or system includes safety by design for children and does not infringe upon their rights. The EU AI Act, UNICEF AI guidance and the UN Convention on the Rights of the Child should form the minimum baseline for protection, with scope for additional safeguards as needed.
Older People
The Committee heard evidence that 41% of people over 75 years of age have never used the internet. Stakeholders noted that increasing digitalisation can cause older people to feel excluded. Age Action highlighted ageist discrimination arising when digitalisation to reduce costs fails to consider the needs of older customers, who may consequently be unable to access services, face higher costs or encounter safeguarding risks.
The report recommends that older people, their families and those providing care for them should be collaborated involved throughout AI product development, including in datasets, product design, testing and rollout. AI should complement rather than replace human contact, with human support options for services being a legal requirement.
The implementation of strategies, including Digital for Care, the HSE data ecosystem roadmap, and the Health Information Bill 2024, should be progressed, with implementation accounting for AI developments and the issues identified in the report.
Disability
Disability The Committee received evidence that 22% of the population declared a disability in the 2022 census, with disabled people's households facing between €488 and €555 in additional weekly costs. A recent report found employment for disabled people in Ireland to be 20% below the EU average, the worst in the European Union.
Witnesses stressed significant risks that AI-based systems pose to equal participation, citing the OECD's 2023 report "Using AI to Support People with Disability in the Labour Market: Opportunities and Challenges". The Committee heard that training data biases mean current AI systems reflect and perpetuate existing inequalities due to the significant underrepresentation of disabled people in their development.
The DPO Network indicated that the high-risk classification under the EU AI Act should automatically apply to areas concerning disability and disability supports, particularly in legal, medical and educational contexts.
The Committee recommends that the National AI Office should have a dedicated role for, or a board including members from, Disabled Persons' Organisations and disability organisations. This would ensure that the benefits and risks posed by AI for people with disabilities remain at the forefront of policy decisions, particularly in the development of ethical regulations.
Resources should be provided to DPOs, disability organisations and disabled people to ensure continued expertise provision and collaboration with Government Departments, State agencies and other bodies. Deaf people should be central to the design, development, and deployment of sign language AI technologies and supported in engaging with national and EU-level frameworks that promote their linguistic autonomy.
AI Literacy and Public Awareness
The Committee recommends a coordinated national effort across education, from primary school to workplace, to support AI literacy. Teachers and workers should understand the opportunities and limitations of AI tools, with equitable access to these tools ensured.
Information for the public should address digital literacy, critical thinking, rights, consent, profiling, bias, scams, bad actors, potential harms and benefits, participation in public policy around AI, engagement with or opting out of AI on devices, ensuring digital rights are respected, and accessing information on how AI systems are used in public service bodies.
The report emphasises that education on AI should not come at the expense of softer skills, including strategising, critical thinking, communication and emotional intelligence. These capabilities are identified as equally important for using technology to maximum advantage in an ethical and responsible manner.
Governance Structures
Beyond the National AI Office, the Committee makes several recommendations regarding advisory and governance bodies. The AI Advisory Council, which currently operates on a voluntary basis, and of which William Fry Partner, Dr. Barry Scannell is a member, should be placed on a permanent footing with State funding and broad societal representation.
An AI Observatory should be established to complement the National AI Office, tracking real-time issues and projecting future impacts on jobs and skills to provide independent forecasting and policy advice.
The Committee recommends that the Government establish a Citizens' Assembly on Artificial Intelligence, Digitalisation and Technology to facilitate inclusive public dialogue and democratic input on AI policy and ethics. The Joint Committee itself should be established on a permanent basis to provide ongoing oversight and guidance on AI policy and regulation.
International Engagement
The Committee addresses the AI Summit planned as part of Ireland's EU Presidency in October 2026. The report recommends that the Summit engage a diverse range of stakeholders and position Ireland as a leader in the global debate on AI adoption, governance, and ethics, with a specific focus on upskilling and reskilling for the AI era.
Conclusion
The First Interim Report represents a comprehensive initial assessment of AI governance challenges and opportunities facing Ireland. The 85 recommendations span institutional architecture, regulatory implementation, sectoral impacts and fundamental rights protection.
The report's emphasis on treating the EU AI Act as a minimum baseline, combined with calls for robust implementation of existing regulations and meaningful stakeholder engagement, signals a direction of travel that prioritises both innovation and protection. The Committee's modular approach means further interim reports will follow, with planned modules addressing the creative industries, AI and the State, and energy considerations in greater depth.
Organisations operating in Ireland should monitor the establishment of the National AI Office and the development of regulatory sandboxes, while ensuring their AI governance frameworks align with the EU AI Act's requirements as they come into effect. The emphasis on procurement processes, algorithmic impact assessments and public sector transparency obligations merits particular attention from entities engaging with Government contracts or deploying AI in public-facing services.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
