Inter-Parliamentary Union

Parliaments worldwide have established a clear and consistent position on AI. In October 2024, the 149th IPU Assembly adopted the resolution The impact of artificial intelligence on democracy, human rights and the rule of law. In November 2025, parliamentary representatives from 65 countries adopted the Kuala Lumpur Declaration on Parliaments and Responsible AI. These commitments stressed that international cooperation must prioritise the protection of human rights; that the concentration of AI power in a few actors is unacceptable; and that the Global South must have an equal voice in shaping the rules. From this perspective, four outcomes would mark a successful first Dialogue. First, a Co-Chairs' summary that names concrete areas of convergence, including on safety, transparency and human oversight, and acknowledges where common understandings remain absent, such as on red lines technology cannot be allowed to cross. Second, a clear pathway to the 2027 Dialogue and the High-Level Review of the Global Digital Compact, so that the work of this session feeds visibly into the international processes parliaments will be asked to scrutinise and implement. Third, treatment of AI capacity-building, particularly for countries and representative institutions in the Global South, as a key deliverable. Parliaments will be responsible for translating international commitments into domestic law and holding governments to account. Fourth, recognition that elected legislatures are part of the international AI governance architecture. Parliaments have the constitutional mandate to legislate on AI, oversee its public deployment, and protect the rights and welfare of the people they represent. The Dialogue's contribution lies in informing parliamentary action at national level by building international consensus and addressing gaps and potential risks in divergent governance approaches. Parliamentary participation throughout would draw on the body of legislative practice now being built, strengthen democratic legitimacy and support national implementation.

• AI capacity-building
• Interoperability of governance approaches
• Protection and promotion of human rights
• Transparency, accountability, and human oversight

These four priorities reflect the areas where the parliamentary mandate is clearest and where international cooperation can most directly support the work parliaments are already undertaking. Human rights lie at the centre of the parliamentary position on AI. Parliaments are passing legislation on data protection, anti-discrimination, online safety and AI-generated harms. The Dialogue can support this work by reaffirming that AI systems are governed by existing international human rights law and by helping align approaches across jurisdictions. Transparency, accountability and human oversight are core constitutional functions of parliament. Parliaments apply them to AI through committee inquiries, public hearings, parliamentary questions and budget scrutiny. International alignment on transparency obligations, incident reporting and meaningful human review of automated decisions would strengthen these tools and the prospects of consistent oversight across borders. AI capacity-building is the test of whether the Dialogue is genuinely universal. Capacity gaps are felt acutely by parliaments themselves across jurisdictions. Most parliaments lack the technical expertise and access to independent advice they need to legislate and oversee AI effectively. Building parliamentary capacity alongside that of governments belongs at the centre of the international agenda. Interoperability of governance approaches matters because parliaments do not legislate on AI in isolation. National frameworks influence one another, but fragmentation across regulatory blocs raises compliance costs, weakens accountability and complicates the parliamentary task of ensuring coherence between domestic legislation, evolving international standards and international human rights law. The Dialogue can help align these core elements while respecting national sovereignty.

Three cross-cutting issues are central to the parliamentary mandate on AI and warrant explicit attention in the Dialogue. First, the impact of AI on democracy. AI-generated disinformation, deepfakes targeting candidates and elected representatives, AI-facilitated gender-based violence - including non-consensual intimate imagery, which disproportionately affects women in public life - and the use of AI in election campaigns are reshaping the conditions in which democratic politics operates. Parliaments have called for urgent action in this area. The role of large language models in mediating how citizens access information, and the political influence exercised by the small number of firms that develop and deploy them, raises further democratic concerns. Parliaments have called for urgent action in this area. Second, the role of parliaments and democratic institutions. The four consolidated clusters address governments and other stakeholders but do not explicitly address parliaments as the institutions that pass legislation, oversee government AI use, scrutinise public expenditure on AI and translate international commitments into binding national law. Recognising parliaments as part of the architecture would strengthen both the democratic legitimacy and the national implementation of the Dialogue's outputs. Third, the rights of children and young people. AI systems are reshaping the environments in which children grow up, learn and exercise their rights. Parliaments are responding with legislation on online harms, age-appropriate design and protection from harmful AI-generated material. The international community has begun to address children's rights and AI through joint statements; the Dialogue should ensure these concerns are reflected in its substantive work, including in the human rights cluster. These issues cut across the four consolidated clusters and would benefit from explicit treatment in the Co-Chairs' summary.

The parliamentary perspective draws on action recorded in over sixty parliaments in different regions by the IPU. Three patterns are visible. On human rights, transparency and oversight, parliaments are confronting AI systems deployed in public administration - for instance, in welfare, immigration, policing and justice - without adequate transparency or appeal mechanisms. Parliamentary committees in different regions are conducting inquiries, but the absence of common international standards on transparency and incident reporting makes cross-border learning difficult. AI-generated harms to women in public life, including women parliamentarians, are now a recognised category of online violence in legislation in a growing number of countries, but responses remain uneven. On capacity, the gap between AI-leading economies and the rest of the world is now a defining feature of the global landscape. The risk is most acute for parliaments in the Global South, which face AI legislation and oversight responsibilities without the technical staff, advisory infrastructure or research services available to their counterparts in major economies. Without targeted investment, the asymmetry between executives and legislatures will widen. The demand for capacity-building, peer exchange and connection to expertise is high but the resources to meet it are limited. On interoperability, there is no shared international point of reference on core questions such as AI red lines or risk thresholds. Parliaments scrutinising government compliance with multiple, non-aligned standards face an important burden. Agreements must be grounded in the UN Charter and international human rights law, in keeping with the framing already adopted in the UNESCO Recommendation on the Ethics of AI, the Global Digital Compact and UNGA Resolution 78/265.

The parliamentary community has committed to engage actively in international processes on AI governance. From that perspective, the Dialogue's distinctive value lies in three roles. First, it is the only universal AI governance platform anchored in the United Nations. Existing initiatives have produced important standards but reach only parts of the international community. The Dialogue gives every Member State a seat at the table and to participate on an equal footing. This is essential to legitimacy and to the parliamentary commitment to an equal voice for the Global South. Second, it can connect science to policy. Through the Independent International Scientific Panel on AI, the Dialogue can ground policy discussions in independent, evidence-based assessments of AI capabilities, risks and impacts. Parliaments depend on this kind of analysis for legislative scrutiny and committee work; few national systems can produce it alone, and parliamentary research services need accessible international references. Third, it can foster coherence across a fragmented landscape. The Dialogue can help align approaches on shared problems such as AI safety and security. This supports the parliamentary task of ensuring coherence between domestic legislation and evolving international standards. To realise these roles, the Dialogue should embed three commitments: meaningful inclusion of parliaments and democratic institutions; attention to the perspectives and needs of Global South; and follow-through, so that outputs influence both national action and corresponding UN processes.

On standards and policy, the Dialogue should adhere to UN General Assembly Resolutions, connect with the OECD AI Principles, UNESCO Recommendation on the Ethics of AI, ITU AI for Good architecture, Council of Europe Framework Convention on AI, and emerging discussions on regional regulatory frameworks. Regional bodies - including the African Union, ASEAN, the Organization of American States and the Pacific Islands Forum - bring contextual perspectives that should be visible in the Dialogue. On engagement with parliaments, the IPU offers an established channel for engaging the world's national parliaments. The mandate adopted by the parliamentary community consists of the IPU resolution "The impact of artificial intelligence on democracy, human rights and the rule of law" (October 2024), the "Kuala Lumpur Declaration on Parliaments and Responsible AI" (November 2025) and the IPU motion on "Advancing parliamentary action on artificial intelligence" (April 2026). The IPU works closely with partners including the Commonwealth Parliamentary Association, UNDP, ITU, the IGF and regional parliamentary networks to support parliamentary action on AI, and support to nationalise emerging international agreements. On evidence, the outputs of the Independent International Scientific Panel should also be designed for use by national legislatures, with summaries adapted for non-specialist audiences. This would meet a clear gap in current parliamentary access to independent international expertise on AI. The Dialogue's added value is to bring these strands together with universal participation, intergovernmental authority and connection to the High-Level Review of the Global Digital Compact.

From a parliamentary perspective, four points would strengthen its inclusiveness and substance. Recognise parliaments as a stakeholder constituency, as they bring the voices of the public they represent. Parliamentarians and senior officials should be visible across the Dialogue: as speakers in thematic clusters, as participants in the multi-stakeholder plenary, and potentially through a dedicated parliamentary segment co-organised with the IPU and partners. The plan to have each thematic cluster co-chaired by a Member State and a relevant stakeholder is a strong feature. Including a parliamentarian as a thematic co-chair in at least one cluster would signal that elected institutions are partners in this process. Ensure regional balance and Global South voices. Member State and stakeholder interventions can be designed so that thematic discussions are not dominated by participants from a small number of countries. Criteria for a regional balance would help give effect to an equal voice for all nations.

Parliamentarians are largely absent from international AI governance forums, despite parliaments being the institutions that pass AI legislation, oversee government AI use, scrutinize public budgets and translate international commitments into national law. Inclusion can be operationalised through a parliamentary segment in the Dialogue programme, parliamentary speakers across thematic clusters, parliamentary participation in national delegations to the high-level segment, and a standing channel to the IPU. The parliamentary community has consistently called for an equal voice for the Global South in shaping international AI governance. The Dialogue can give effect to this through attention to regional balance in thematic discussions and panel composition, and practical measures such as multilingual interpretation and hybrid participation. Parliaments hear directly from these communities most affected by AI harms - workers facing rapid automation; women confronting AI-facilitated gender-based violence; people subject to algorithmic decisions in welfare, immigration and justice systems - through constituency work, hearings and inquiries, and represent them in the legislative process. Structured access for parliamentarians, alongside civil society organisations, would help bring these voices into the Dialogue.

Structured peer exchange between governance actors. Format thematic discussions around concrete case studies raised by national legislators and other stakeholders. Live engagement with the Scientific Panel. The presentation of the Panel's first report should include pre-submitted questions from Member States and stakeholders. It can include panels with practitioners to help translate scientific findings into policy-relevant discussion. The Panel's outputs should be made available in formats accessible to parliamentary research services and committees. Side-event slots can be used to host focused, action-oriented sessions on themes that warrant deeper treatment than may be allowed in the plenary.

Parliamentary action across core lawmaking, oversight and representation functions. The IPU has tracked action by parliaments in different regions, including passing legislation related to AI, establishing dedicated parliamentary committees and cross-party groups on AI, conducting public and stakeholder hearings, conducted inquiries into the adoption and impact of AI, and scrutinizing public-sector use of AI. These examples illustrate the role parliaments play in giving international standards practical effect at national level; and the importance of parliaments as the bridge between the public and international agreements on AI. Inter-parliamentary peer exchange and dialogue. The IPU, international partners and regional parliamentary networks have developed peer-learning approaches the Dialogue could draw on. The first international parliamentary conference on AI, held in Kuala Lumpur in November 2025, brought together parliamentary representatives from 65 countries and produced the body of recommendations in the "Kuala Lumpur Declaration on Parliaments and Responsible AI" that now guides the parliamentary community's engagement on AI.