Institutional aspects of Artificial Intelligence in the context of the European integration
The Committee on Constitutional Affairs adopted this own-initiative report on 1 October 2026 by 21 votes to 3, with no abstentions. It proposes that Parliament set out a stronger institutional approach to artificial intelligence (AI), combining effective implementation of the Artificial Intelligence Act (AI Act), democratic oversight, human responsibility and European technological sovereignty. The report draws on opinions from the Committee on the Internal Market and Consumer Protection and the Committee on Civil Liberties, Justice and Home Affairs.
Human-centred governance and fundamental rights
The report proposes that Parliament affirm that AI governance is a constitutional and strategic issue, not only a technical regulatory matter. It recommends that Parliament insist that AI must not weaken the rule of law, public security, democratic accountability or Parliament’s prerogatives as co-legislator and budgetary authority.
It proposes that Parliament uphold a “human decision-maker” principle: AI may support institutional work, but must not replace elected representatives’ mandates, political responsibility, judicial judgment or meaningful human oversight. AI used by EU institutions and public administrations should be transparent, accountable, auditable and subject to effective redress, including human review of automated decisions.
The report recommends that Parliament call for an amendment to the Charter of Fundamental Rights of the European Union to include rights linked to widespread AI use. These would include protection of personal identity and likeness against deepfakes, algorithmic transparency, digital self-determination, protection against algorithmic discrimination, and accountability and liability for harm.
It also proposes strict safeguards for AI in high-risk fields, including employment, law enforcement and access to public services, with possible prohibitions where fundamental rights cannot otherwise be protected. AI should remain an auxiliary tool in legal work and must not take part in judicial decision-making.
European capacity and coherent implementation
The report proposes that Parliament support European AI capabilities, secure and sovereign cloud infrastructure, access to high-quality data and advanced computing capacity, and reduced dependence on non-EU providers. It recommends prioritising AI tools developed, operated and hosted in the EU, particularly for use by EU institutions and for sensitive data.
It calls for a genuine internal market for AI, addressing disparities in digital infrastructure, data capacity and expertise between Member States. The report recommends uniform, proportionate and innovation-friendly implementation of the AI Act and the wider EU digital framework, while avoiding unnecessary administrative burdens, duplicate reporting requirements and national gold-plating that could fragment the market or hinder small and medium-sized enterprises and start-ups.
The committee recommends that Parliament call on the Commission to provide the European AI Office with stable financial, technical and human resources, including specialised expertise. It also proposes closer coordination between the AI Office, the European Artificial Intelligence Board, the European Data Protection Board, the European Data Protection Supervisor, other EU bodies and national authorities.
Parliamentary capacity and interinstitutional oversight
The report proposes that Parliament consider creating a dedicated administrative unit, potentially called the European Parliament AI Observatory. It would build on and complement existing services, including the European Parliamentary Research Service and the Panel for the Future of Science and Technology, without duplicating structures and in a budget-neutral manner.
The proposed Observatory would support all parliamentary committees, strengthen AI literacy and technical expertise, monitor institutional AI governance, and assess AI’s effects on the EU’s institutional balance and democratic processes. It would support ex post evaluation of AI-related legislation, identify legal gaps, overlaps and inconsistencies, and report regularly to the competent committees, especially the Committee on Constitutional Affairs.
The report recommends structured, regular and transparent coordination between Parliament, the Commission and the Council, potentially formalised through an interinstitutional agreement. It also calls for regular exchanges with national parliaments, national competent authorities and candidate countries. If established, the Observatory should be invited to meetings of the European Artificial Intelligence Board.
Democracy, elections and public participation
The report proposes systematic parliamentary assessment of AI-related risks and opportunities for democratic processes. It highlights risks from disinformation, foreign information manipulation and interference, deepfakes, synthetic media, microtargeting and voter manipulation. It recommends enforcing AI Act transparency obligations for AI-generated content, protecting Parliament’s content and Members against false attribution and deepfakes, and considering watermarking and ownership rules.
It proposes that Parliament call for accelerated reform of EU electoral law to address AI in electoral campaigns, political advertising and voter targeting. It also recommends measures to protect media pluralism, including use of the Digital Services Act and Digital Markets Act.
At the same time, the report recognises that civic AI could improve multilingual citizen participation, access to information and analysis of public contributions, provided its use is transparent, inclusive, accountable and connected to formal decision-making.
International governance and wider impacts
The report recommends that Parliament support a stronger EU role in global AI governance, including participation in multilateral forums such as the United Nations “AI for Good” summit. It proposes that the EU promote a rules-based, human-rights-based approach grounded in democracy, the rule of law and fundamental rights.
It suggests an international AI convention or treaty, with effective global oversight and provisions on the non-proliferation of high-risk or harmful AI technologies. The report also calls for international cooperation to address systemic risks, cybersecurity, digital extractivism, concentration of power, labour-market disruption, inequalities and AI’s growing environmental footprint.
This summary was AI-generated and human-reviewed.