Trustworthy AI for Citizen Participation Creating conditions in which artificial intelligence responsibly supports accessible, meaningful, and impactful citizen participation in policymaking. This plan renders recommendations set forth in the OECD report Artificial Intelligence and the Future of Citizen Participation: Typology of Applications, Opportunities and Challenges for Democratic Innovation, published in 2026. ^^ The report draws upon desk research and analysis of 50 uses of artificial intelligence in citizen participation processes across 22 OECD member and partner countries. It identifies opportunities for AI to improve the design, accessibility, scale, communication, and resource efficiency of participation processes while warning that AI adoption entails ethical, operational, exclusion, public-resistance, and inaction risks. ^^ The OECD recommends that governments establish guardrails, strengthen institutional and technical enablers, engage citizens and stakeholders in shaping and governing AI systems, and continue improving the institutional design and accessibility of citizen participation independently of technology. ^^ The report rejects a purely technological or techno-solutionist approach. AI tools should complement rather than replace human participation, judgment, deliberation, and accountability. ^^ Source: https://doi.org/10.1787/a1ee2e0a-en ^^ This is an adaptation of an original work by the OECD. The opinions expressed and arguments employed in this adaptation should not be reported as representing the official views of the OECD or of its Member countries. ^^ Submitter's Note: It has been compiled from the source PDF and rendered in StratML format by ChatGPT. It has been reviewed in the form at https://stratml.us/forms/Claude/Part1.html Organisation for Economic Co-operation and Development OECD 9f297adb-1dc6-44db-bd17-ac3927dcaa58 Produced the report and recommendations concerning trustworthy uses of artificial intelligence in citizen participation. OECD Public Governance Directorate Prepared the report and supports governments in strengthening public governance, open government, citizen participation, and trustworthy adoption of artificial intelligence. OECD Public Governance Committee Approved and declassified the report on June 22, 2026. OECD Working Party on Open Government Provided consultation through its delegates during preparation of the report. Bertelsmann Stiftung Collaborated in initiating and conceptualizing the report and provided financial support, expertise, and recommendations. Giulia Cibrario OECD policy analyst who drafted the report. Elsa Pilichowski Director of the OECD Public Governance Directorate under whose leadership the report was prepared. Nejla Saula Provided strategic direction as Acting Head of the OECD Anti-Corruption, Integrity and Open Government Division. David Goessmann Provided strategic direction as Head of the OECD Open Governance Unit. Mauricio Mejia Galvan Supported conceptualization of the report. Raphaël Pouyé Provided strategic comments throughout development of the report. Dominik Hierlemann Senior Advisor at Bertelsmann Stiftung who provided insights concerning the content, structure, and recommendations of the report. Angela Jain Senior Project Manager at Bertelsmann Stiftung who provided insights concerning the content, structure, and recommendations of the report. Stefan Roch Project Manager at Bertelsmann Stiftung who provided insights concerning the content, structure, and recommendations of the report. Artificial intelligence strengthens meaningful, inclusive, trustworthy, and consequential citizen participation in policymaking. 4d5042cf-4dcc-45a7-950d-f022e6b765fc To guide governments in adopting fit-for-purpose AI tools that expand political agency, improve public participation, protect civic space, and preserve human judgment and accountability. b006babf-48bb-46d0-bb3f-9a699bdf02b3 Democracy Align AI development and use with democratic values, human rights, the rule of law, and meaningful public agency. Human Agency Use AI to support rather than replace citizens, practitioners, public officials, and accountable decision makers. Fairness Prevent skewed data, discriminatory outcomes, unequal access, and exclusion of underheard communities. Transparency Disclose how AI systems operate, where they are used, what limitations they have, and how their outputs affect participation. Privacy Safeguard personal information, intellectual property, security, and responsible data governance throughout the AI lifecycle. Accessibility Provide participation opportunities that accommodate differences in language, ability, digital access, literacy, and preferred means of engagement. Accountability Maintain human responsibility, auditability, oversight, contestability, and effective mechanisms for redress. Openness Promote open standards, interoperability, accessible documentation, and appropriately open software, models, and public infrastructure. Evidence Evaluate risks, impacts, participation quality, and public outcomes throughout the AI lifecycle. Subsidiarity Use AI as a complementary aid to human activity and apply it only when it adds value to participation. Guardrails Protect democratic values, civic space, fairness, privacy, security, transparency, inclusion, and public trust when AI is used for citizen participation. 07b33fa5-6d0d-4dee-853f-b45cec5187d0 1 Governments Establish legal, ethical, institutional, technical, and procedural safeguards for AI-supported participation. Citizens Participate in public decision-making and depend upon accessible, fair, transparent, and trustworthy processes. Public Officials Select, procure, operate, oversee, and remain accountable for AI systems used in participation processes. Civil Society Organizations Protect civic space, advocate for affected communities, scrutinize AI uses, and hold institutions accountable. AI Providers Develop and supply AI systems whose design, data, documentation, security, and performance affect public participation. The OECD Framework for Trustworthy AI in Government identifies guardrails as measures for anticipating and managing risks. In the context of citizen participation, those risks include skewed data, discriminatory or inaccurate outcomes, surveillance and other threats to civic space, lack of transparency and explainability, weakened deliberation, overreliance on AI, privacy and cybersecurity failures, exclusion, public mistrust, and inadequate means of contesting harmful outcomes. Democratic Alignment Ensure that AI systems and associated governance processes comply with human rights, democratic values, civil freedoms, and the rule of law. 76d3bd10-1865-436e-98bb-aacc85e95e97 1.1 Legislators Establish laws governing responsible and rights-respecting uses of AI. Regulators Develop, apply, and enforce regulations and safeguards governing AI systems and public participation. Human Rights Organizations Assess and challenge AI practices that may infringe civil freedoms, equal treatment, privacy, or political agency. Governments should choose and govern AI tools consistently with the OECD AI Principles, particularly respect for human rights, democratic values, fairness, privacy, transparency, explainability, robustness, security, safety, and accountability. Particular caution is warranted where AI could enable surveillance, manipulate information, constrain expression, or otherwise undermine civic space. Data Quality Improve the quality, relevance, representativeness, governance, and auditability of data used by AI systems. acb9fec0-4135-4ac0-957b-77a1ce495fac 1.2 Data Stewards Manage data quality, provenance, access, security, retention, documentation, and lawful use. AI Auditors Assess data, models, outputs, controls, and impacts throughout the AI lifecycle. Skewed, incomplete, outdated, or unrepresentative data can produce inaccurate or adverse outcomes and systematically underrepresent communities. The OECD recommends investment in high-quality data, AI audits, lifecycle risk assessment, and measures to close data divides, including those affecting low-resource languages. Privacy & Security Protect personal information, intellectual property, system security, and public participation data throughout the AI lifecycle. fdc1a6ff-0fd4-49b2-9b29-10e647971f51 1.3 Privacy Officials Establish and oversee privacy-preserving practices for collection, processing, sharing, retention, and deletion of data. Cybersecurity Officials Protect AI systems, participation platforms, data, identities, and communications from unauthorized access, manipulation, and disruption. Governments should establish clear frameworks, guidance, safeguards, and accountability mechanisms for privacy-preserving processing of personal data. Robust data governance should address privacy, security, intellectual property, access controls, provenance, retention, and authorized reuse. Transparency Disclose the existence, purposes, operation, limitations, data dependencies, and public effects of AI systems used in participation. 4fb77e95-6e82-486c-a9a8-aa794f749743 1.4 Participation Practitioners Explain how AI is used in participation processes and how citizens’ contributions are processed, summarized, moderated, or interpreted. Auditors Examine system documentation, registries, decisions, outputs, and claims concerning AI performance and accountability. The OECD recommends mapping existing practices, publishing registries of automated systems, explaining how AI systems work and are used, and considering publication of source code for government-developed systems when appropriate. Transparency should be meaningful to citizens rather than limited to technical disclosure. Openness & Interoperability Use open standards, interoperable infrastructure, and appropriately open software, documentation, and model components when they improve accountability and public value. 0da51812-96f9-47ff-94cf-693e1ffb03dd 1.5 Standards Developers Develop and maintain transparent technical specifications supporting portability, interoperability, auditability, and reuse. Civic Technology Developers Build transparent and interoperable tools for public information, consultation, deliberation, and engagement. Open-source software has played an important role in civic technology, but openness in AI may involve several distinct layers, including source code, documentation, training data, and model weights. Governments should avoid misleading claims of openness and consider open standards, open weights, and open-source components according to context, risks, feasibility, and public value. Inclusion Prevent AI-supported participation from excluding people because of digital access, language, disability, literacy, socioeconomic circumstances, or other barriers. 70ebf21c-2b48-4bac-a284-bca0b3f4dc82 1.6 Underheard Communities People and communities whose views, needs, languages, or circumstances are often inadequately reflected in public participation. People with Limited Digital Access Citizens who require in-person, telephone, paper-based, assisted, or other low-technology means of participation. Accessibility Specialists Help design participation processes and tools usable by people with differing abilities, languages, devices, and levels of literacy. Governments should complement digital processes with in-person and low-technology alternatives, proactively involve distant and underheard groups, close language and data divides, and use efficiency gains from AI to improve resource allocation toward inclusive participation channels. Human Oversight Keep AI systems subsidiary and complementary to human participation, deliberation, judgment, facilitation, and accountable decision-making. d7d25777-5cc4-4879-b54b-b40c47743ab1 1.7 Decision Makers Retain responsibility for interpreting evidence, weighing trade-offs, explaining decisions, and responding to public input. Facilitators Preserve the quality, depth, fairness, and human character of dialogue and deliberation. AI should provide citizens with more and better opportunities to participate rather than substitute predictions, simulations, synthetic views, or automated summaries for genuine consultation. Governments should guard against overreliance, automation bias, hallucinations, loss of context, and deterioration in the depth and quality of deliberation. Accountability & Redress Establish clear responsibility, oversight, contestability, grievance, correction, and redress mechanisms for AI-supported participation. 45c2d833-1d11-4f0a-b717-be19d1c85fee 1.8 Oversight Bodies Review AI systems, participation processes, complaints, impacts, compliance, and corrective actions. Participants Require accessible means to question, correct, appeal, or challenge AI-supported processing and its effects. Governments should provide accessible information about AI uses and establish strong accountability and redress mechanisms. Citizens should be able to identify who is responsible, understand how their contributions were handled, report problems, request corrections, contest consequential uses, and obtain meaningful responses. Capacity Strengthen the institutional, human, technical, financial, and procurement capabilities required to adopt AI responsibly in citizen participation processes. 7aa3c883-8b34-497e-b03d-5d321fb191bb 2 Governments Develop coordinated capabilities, infrastructure, policies, resources, and operating practices for trustworthy AI-supported participation. Public Officials Develop the knowledge and judgment needed to select, use, oversee, and evaluate AI tools responsibly. Participation Practitioners Integrate appropriate AI tools into the design, facilitation, analysis, communication, and evaluation of participation processes. Public Procurement Officials Acquire interoperable, secure, transparent, fit-for-purpose, and sustainable AI systems under appropriate contractual terms. Public Technology Officials Develop and maintain the technical infrastructure, data, security, interoperability, and support services needed for responsible AI adoption. Researchers Generate evidence about the capabilities, limitations, risks, impacts, and appropriate uses of AI in citizen participation. The OECD identifies organisational and institutional barriers that may prevent governments from realizing the potential benefits of AI. These include gaps in awareness, skills, literacy, investment, infrastructure, procurement capacity, coordination, and the ability to move beyond isolated experiments. ^^ Capacity should not be understood solely as technical proficiency. Public officials and participation practitioners also need critical judgment concerning when AI is appropriate, how its outputs should be interpreted, what risks it creates, and when human methods or non-digital alternatives are preferable. Awareness Increase understanding among public officials and participation practitioners of the opportunities, limitations, risks, and appropriate uses of AI. 8b310591-ddd8-4d28-826d-540db27a4f29 2.1 Government Executives Promote informed consideration of AI opportunities and risks across public institutions. Participation Managers Identify where AI may improve participation and where technological intervention may be unnecessary or harmful. Governments should raise awareness of existing applications, lessons, safeguards, and limitations. Greater awareness can help institutions avoid both indiscriminate adoption and missed opportunities arising from unfamiliarity with relevant tools. Literacy Develop practical and critical AI literacy among public officials, practitioners, participants, and affected communities. b11d71e0-6ee8-454c-9ce1-ce5705e414d0 2.2 Civil Servants Learn to use, assess, question, document, and supervise AI systems in the performance of public duties. Citizens Develop sufficient understanding to participate knowingly in processes involving AI and to identify limitations or potentially harmful uses. Training Providers Deliver multidisciplinary learning opportunities covering technical, ethical, legal, operational, and democratic implications. Training should extend beyond operational instruction. It should promote critical understanding of automation bias, hallucinations, data limitations, uncertainty, privacy, security, fairness, explainability, and the effects of AI on participation and deliberation. Expertise Build multidisciplinary teams capable of designing, operating, evaluating, and governing AI-supported participation processes. 1aed9584-d8f7-4037-af5c-f7dcbeed0eb6 2.3 AI Specialists Provide technical knowledge concerning models, data, system performance, testing, security, and limitations. Participation Specialists Provide expertise concerning process design, recruitment, accessibility, facilitation, deliberation, communication, and follow-through. Legal & Ethics Specialists Assess compliance, rights, fairness, accountability, privacy, civic-space, and public-interest implications. Social Scientists Evaluate human behavior, institutional effects, social inequalities, public perceptions, and participation outcomes. Responsible adoption requires collaboration among technical experts, participation practitioners, social scientists, legal and ethics specialists, accessibility experts, data stewards, cybersecurity personnel, and affected communities. No single professional discipline can adequately assess all relevant consequences. Infrastructure Develop secure, interoperable, reusable, and accessible digital public infrastructure supporting AI-enabled participation. 6c712ddd-489c-47cd-bf38-06193406ab5b 2.4 Digital Government Authorities Coordinate shared platforms, standards, identity services, data resources, security controls, and reusable public technology. Platform Operators Operate reliable and accessible systems through which citizens obtain information, contribute views, deliberate, and receive feedback. Standards Developers Enable interoperability, portability, auditability, reuse, and avoidance of proprietary lock-in. Investment in digital public infrastructure can reduce duplication, support interoperability, improve security, and enable public institutions to reuse trustworthy components. Infrastructure should support both front-office interactions with citizens and back-office government operations. Investment Provide sustained and proportionate resources for the responsible design, testing, operation, evaluation, maintenance, and improvement of AI-supported participation. b655ca5c-9288-486d-bdb8-dc5059d4d159 2.5 Budget Officials Allocate resources according to public needs, expected value, lifecycle costs, risks, and evidence of effectiveness. Program Managers Plan and manage the human, financial, technical, and organisational resources required for sustainable implementation. AI adoption requires more than initial funding for a pilot or software licence. Governments should account for data preparation, integration, accessibility, security, training, human oversight, evaluation, maintenance, updates, redress, and non-digital participation alternatives throughout the system lifecycle. Experimentation Test AI applications in controlled settings while documenting assumptions, risks, results, failures, and lessons. 1976b625-cf54-45f2-9aaf-8c22112ed346 2.6 Public Innovation Teams Conduct responsible experiments and connect technical testing with participation needs and public outcomes. Pilot Participants Provide informed feedback concerning usability, accessibility, trust, fairness, and effects on participation quality. Evaluators Assess whether experiments generate reliable evidence of value, risk, scalability, and transferability. Pilots can support learning under uncertainty, but experimentation should not expose citizens to unmanaged risks or become a substitute for institutional adoption and evaluation. Experiments should include clear purposes, hypotheses, safeguards, success criteria, documentation, and decisions concerning continuation, modification, scaling, or termination. Coordination Coordinate AI-supported participation strategies, tools, standards, knowledge, and responsibilities across levels and branches of government. b158e748-b969-469c-b5c9-f13fe1f7b990 2.7 National Governments Provide enabling strategies, common resources, standards, guidance, and support for public institutions. Regional Governments Adapt and coordinate capabilities and participation practices across regional institutions and communities. Local Governments Develop and apply AI tools in proximity to citizens while sharing lessons and reusable practices. Most government adoption remains ad hoc or pilot-based, although local governments are prominent sources of experimentation. Coordinated approaches can reduce duplication, connect isolated initiatives, disseminate lessons, and make trustworthy capabilities available to institutions that lack resources to develop them independently. Scaling Expand successful AI applications when evidence demonstrates that they are trustworthy, effective, reusable, and appropriate to additional contexts. 48510323-eede-418a-bbbf-59066b84e853 2.8 Program Sponsors Authorize expansion based on evidence, safeguards, resource requirements, public value, and readiness. Implementation Teams Adapt proven tools and practices while accounting for differences in institutions, populations, languages, and policy contexts. Scaling should follow evidence rather than technological enthusiasm. Governments should distinguish successful demonstrations from systems ready for broader use and should preserve evaluation, human oversight, accessibility, security, and accountability as use expands. Procurement Acquire AI systems through processes that promote fitness for purpose, transparency, interoperability, competition, accountability, and long-term public value. 204a078f-ae23-4017-b0d2-0297440c6c32 2.9 Procurement Authorities Specify requirements, evaluate suppliers, negotiate safeguards, and manage contractual performance and risk. Technology Suppliers Provide systems, documentation, support, testing evidence, security assurances, interoperability, and contractual accountability. Competition Authorities Promote competitive markets and scrutinize practices that create avoidable dependency or constrain public choice. Procurement should consider data and vendor lock-in, interoperability, portability, audit rights, access to documentation, data ownership, privacy, security, model updates, performance monitoring, accessibility, redress, termination, and migration to alternative systems. Governments should acquire technology to meet clearly defined participation needs rather than redesign participation around a supplier’s product. Evaluation Assess AI-supported participation against clearly defined purposes, risks, costs, process quality, inclusion, public value, and policy outcomes. db31c6d1-6c75-46eb-bab8-18b4f175195f 2.10 Independent Evaluators Provide credible assessments of system performance, participation quality, social effects, costs, risks, and outcomes. Participants Provide evidence concerning accessibility, usability, fairness, trust, understanding, influence, and satisfaction. Decision Makers Use evaluation findings to continue, revise, expand, restrict, or discontinue AI applications. Evaluation should examine more than speed, scale, or administrative savings. It should determine whether AI improves the accessibility, representativeness, depth, quality, transparency, consequentiality, and legitimacy of participation and whether claimed benefits exceed financial, institutional, democratic, and opportunity costs. Citizen Partnership Engage citizens and stakeholders in shaping the development, deployment, governance, and evaluation of AI systems affecting public life. 0fe290d8-da39-48c6-b4cc-bb14847b02ec 3 Citizens Contribute their needs, values, knowledge, experiences, concerns, and judgments to decisions concerning AI systems. Governments Create meaningful opportunities for citizens and stakeholders to shape the development, use, regulation, and oversight of AI. Civil Society Organizations Convene communities, articulate public concerns, represent diverse interests, scrutinize AI systems, and support public accountability. Underheard Communities Contribute perspectives and experiences that may otherwise be absent from AI design, data, governance, and evaluation. AI Developers Engage affected communities and incorporate public needs, values, and evidence into AI system design and operation. Participation Practitioners Design and facilitate inclusive processes through which citizens can influence AI-related decisions. Researchers Study participatory approaches to AI governance and evaluate their effects on systems, institutions, citizens, and public outcomes. The report characterizes the relationship between AI and citizen participation as a two-way street. AI tools may support participation, while citizen participation processes may enable the public to shape the design, deployment, regulation, and monitoring of AI systems. ^^ Citizens should not be treated merely as users, data subjects, customers, or recipients of AI-enabled public services. They should be engaged as partners whose knowledge, interests, rights, values, and lived experiences inform decisions throughout the AI lifecycle. ^^ Participation should be consequential rather than symbolic. Governments should clearly identify the issues open to influence, provide relevant information, explain institutional constraints, document how public contributions are considered, and report resulting decisions and actions. AI Governance Use citizen participation processes to involve the public in developing policies, rules, priorities, and safeguards governing AI systems. 583bf487-e05b-4a21-b73c-b45919933740 3.1 Policymakers Invite and consider public input when establishing AI strategies, laws, regulations, standards, and oversight arrangements. Regulators Engage affected stakeholders in identifying risks, developing requirements, assessing impacts, and improving regulatory practices. Citizens Express public values, priorities, expectations, concerns, and acceptable boundaries for AI development and use. Citizen participation can help governments identify societal expectations and trade-offs that cannot be resolved solely through technical expertise. Participation may support the development of AI strategies, ethical principles, legislation, regulation, impact-assessment methods, procurement requirements, and oversight mechanisms. Co-Design Engage intended users and affected communities in defining needs and designing AI systems, services, interfaces, and participation processes. 4e472d9f-fe12-48be-ad4a-3139a340b2c0 3.2 Intended Users Help define needs, requirements, usability expectations, accessibility features, and appropriate uses of AI systems. Affected Communities Identify foreseeable effects, risks, barriers, and unintended consequences based on lived experience. Service Designers Translate public needs and participation findings into system and service requirements. AI Developers Incorporate user and community input into technical design, testing, documentation, deployment, and improvement. Co-design should begin with the public need or participation challenge rather than with a predetermined technology. It can help determine whether AI is appropriate, what functions it should perform, what data it may use, what safeguards are necessary, and what non-AI alternatives should remain available. Deliberation Enable informed public deliberation about the purposes, trade-offs, uncertainties, and societal consequences of AI systems. 07f3f9df-c634-4370-8403-0bf90865ff0e 3.3 Deliberative Participants Consider evidence, exchange perspectives, weigh trade-offs, and formulate informed judgments and recommendations. Facilitators Support balanced, respectful, informed, and inclusive consideration of complex AI issues. Subject-Matter Experts Provide accessible and balanced evidence concerning technical capabilities, limitations, risks, and policy options. Decision Makers Consider deliberative findings and explain their influence on resulting decisions. Representative deliberative processes may be particularly useful for complex AI policy questions involving uncertainty, competing values, and long-term consequences. Participants should receive balanced evidence, sufficient time, skilled facilitation, and opportunities to question experts and one another before reaching recommendations. Lifecycle Participation Involve citizens and affected stakeholders throughout AI system design, testing, deployment, monitoring, evaluation, and retirement. c8621765-ac23-443d-9851-04811f36782d 3.4 System Owners Establish recurring opportunities for public input and response throughout the system lifecycle. System Users Report operational limitations, usability problems, errors, and opportunities for improvement. Affected Persons Provide evidence concerning system effects, harms, exclusions, and unintended consequences. Oversight Bodies Use stakeholder evidence to assess performance, compliance, impacts, and the need for corrective action. Participation should not end when a system is procured or deployed. Continuing engagement can reveal changing needs, emerging risks, performance failures, discriminatory effects, public resistance, and differences between intended and actual uses. Findings should inform modification, restriction, suspension, replacement, or retirement decisions. Public Feedback Provide accessible channels through which people can report experiences, question outputs, identify harms, and propose improvements to AI systems. 42069255-8571-4ab0-aec2-b8d654e33721 3.5 Participants Provide feedback concerning the use and effects of AI in participation processes. Public Service Users Report whether AI-enabled services are understandable, accessible, accurate, fair, and responsive to their needs. Complaint & Redress Officials Receive, investigate, respond to, and resolve reports concerning AI systems and their effects. Feedback mechanisms should be visible, understandable, accessible, and connected to responsible officials. Governments should distinguish general feedback from formal complaints and appeals while ensuring that both can lead to review, correction, learning, and institutional response. Innovation Ecosystems Strengthen networks through which governments, citizens, civil society, researchers, and technology developers co-create trustworthy AI applications for participation. 0a00184e-2dae-4937-af32-75dc0e30e3eb 3.6 Public Institutions Define public needs, provide enabling resources, share lessons, and support trustworthy experimentation and adoption. Civic Technology Organizations Develop participation tools and connect technical innovation with democratic practices and community needs. Academic Institutions Conduct research, support experimentation, evaluate impacts, and contribute multidisciplinary expertise. Technology Enterprises Develop and adapt tools while meeting public-interest requirements for transparency, accessibility, accountability, and interoperability. Community Organizations Connect innovation activities with local knowledge, public needs, underheard groups, and lived experience. The OECD recommends strengthening innovation ecosystems to support co-creation. Such ecosystems can combine public-sector knowledge, technical capability, participation expertise, academic research, civic innovation, and community experience while reducing dependence on isolated institutions or dominant suppliers. Co-Creation Platforms Establish shared environments in which stakeholders can identify needs, develop prototypes, test safeguards, exchange evidence, and improve AI applications. 6030bd9f-27f1-4ca2-b285-e230da4b98ba 3.7 Platform Conveners Provide governance, infrastructure, facilitation, documentation, and access for collaborative development and testing. Public Innovators Develop and test solutions addressing clearly defined public and participation needs. Community Participants Contribute practical knowledge, evaluate prototypes, identify barriers, and judge whether proposed tools address genuine needs. Co-creation platforms may include public innovation laboratories, regulatory sandboxes, civic technology programs, participatory design environments, shared testing facilities, and open digital platforms. Their governance should clarify ownership, decision rights, data use, intellectual property, safeguards, evaluation, and pathways to adoption. Communities of Practice Support recurring peer networks that exchange knowledge, tools, evidence, standards, safeguards, and lessons concerning AI-supported participation. b87ea78e-bd9d-48d6-bf07-52b0fadf2f37 3.8 Public Participation Practitioners Share methods, experiences, needs, failures, and lessons from the use of AI in participation processes. Digital Government Practitioners Share technical resources, governance practices, procurement approaches, security controls, and reusable infrastructure. Researchers Contribute evidence, methods, independent assessment, and knowledge of emerging developments. Civil Society Practitioners Bring public-interest, community, rights, inclusion, and accountability perspectives into professional learning networks. Communities of practice can reduce fragmented learning and repeated mistakes by enabling practitioners across institutions and levels of government to share documentation, tested components, evaluation findings, risk controls, procurement language, and examples of unsuccessful as well as successful applications. Public Communication Communicate clearly about how citizen contributions shape AI policies, systems, safeguards, and institutional decisions. cd29fb31-df4a-4368-a1df-e61136ebda61 3.9 Government Communicators Provide accessible information about participation opportunities, AI uses, public contributions, decisions, and follow-up actions. Process Sponsors Explain the scope of public influence, institutional constraints, decisions reached, and reasons for accepting or declining recommendations. Participants Assess whether the government has accurately represented and meaningfully addressed their contributions. Closing the feedback loop is essential to consequential participation and public trust. Governments should explain what they heard, how contributions were analyzed, what recommendations were accepted or rejected, why decisions were made, who is responsible for implementation, and how progress can be followed. International Cooperation Exchange knowledge and coordinate participatory approaches to AI governance across jurisdictions and international institutions. b7ab7f9d-d03a-4e6f-afb5-eeae0950d747 3.10 National Governments Share practices, evidence, standards, and lessons concerning public engagement in AI governance. International Organizations Convene jurisdictions, develop shared guidance, compare experience, and support cross-border learning. Transnational Civil Society Networks Connect public-interest perspectives and affected communities across national boundaries. AI systems and suppliers often operate across jurisdictions, while their societal implications extend beyond national boundaries. International cooperation can promote shared learning, interoperable safeguards, broader participation, and consideration of communities that may be affected by decisions made elsewhere. Meaningful Participation Improve the quality, accessibility, effectiveness, and public impact of citizen participation while using AI only where it demonstrably adds value. d61b9d50-a6bb-4202-a00b-2b39e1af8872 4 Governments Design participation processes that are consequential, inclusive, transparent, well-resourced, and connected to public decision-making. Citizens Contribute knowledge, preferences, values, experiences, and judgments that improve public decisions and strengthen democratic legitimacy. Participation Practitioners Design and implement participation processes that make effective use of both human capabilities and appropriate technologies. Decision Makers Consider public contributions in good faith and explain resulting decisions and actions. The OECD concludes that AI alone cannot solve the institutional and organizational shortcomings of citizen participation. Governments should continue improving participation processes regardless of whether AI is used. ^^ AI should therefore be viewed as one tool among many that may improve participation under appropriate circumstances. Governments should first define the public purpose, participation objectives, institutional context, and decision-making process before determining whether AI offers meaningful advantages over conventional approaches. ^^ The report explicitly rejects a techno-solutionist approach. Trustworthy AI adoption and meaningful participation should advance together rather than independently. Purpose Define clear public purposes, participation objectives, decision contexts, and expected public value before selecting AI tools. 8f0ce26e-8fc4-4d59-b327-37d0d5cbf0d4 4.1 Program Sponsors Define why participation is being undertaken and what decisions may be influenced. Participation Designers Align participation methods and technologies with clearly defined objectives. Technology selection should follow rather than determine public objectives. Governments should first identify the participation challenge to be addressed, then determine whether AI provides meaningful advantages relative to conventional approaches. Accessibility Design participation opportunities that are understandable, convenient, multilingual, inclusive, and usable by diverse populations. ef9b62bd-79f8-40e5-96e5-c0eec35eb5cf 4.2 Accessibility Specialists Ensure participation processes accommodate diverse abilities, languages, technologies, and circumstances. Community Organizations Help identify and reduce barriers affecting participation by underrepresented groups. AI can improve accessibility through translation, transcription, conversational assistance, and adaptive interfaces. However, accessibility also depends upon organizational design, outreach, scheduling, language, trust, institutional culture, and continued availability of non-digital participation methods. Scale Expand opportunities for meaningful participation while preserving deliberative quality, representativeness, and public trust. 35f5f7e3-c6e0-4fc2-bcfd-8aef66547289 4.3 Participation Managers Balance broader participation with adequate opportunities for learning, dialogue, reflection, and informed judgment. Facilitators Preserve constructive interaction and deliberative quality as participation expands. One of AI's greatest opportunities is enabling participation at scales previously impractical. However, increasing the number of participants should not diminish the depth, quality, or representativeness of deliberation or weaken participants' opportunities to understand one another. Communication Improve public understanding of participation opportunities, processes, findings, and resulting governmental actions. 2b0f00b3-79f3-4aef-8ee4-5d3d2748d220 4.4 Government Communicators Provide timely, understandable, accurate, and balanced information before, during, and after participation processes. Media Organizations Inform broader public discussion concerning participation processes and resulting decisions. AI may assist governments in preparing summaries, translating materials, tailoring communications, and helping citizens navigate participation opportunities. Human review should ensure that communications remain accurate, balanced, and faithful to participants' contributions. Consequentiality Ensure citizen participation meaningfully influences public decisions and is not merely consultative or symbolic. bc3a84d5-4ca4-4620-a871-d780c3b8a814 4.5 Decision Makers Consider public recommendations seriously and explain resulting decisions. Participants Provide informed recommendations with the expectation that they will receive meaningful consideration. The report identifies disconnection from decision-making as one of the principal weaknesses of participation. Governments should clearly explain how citizen input affects policy development, implementation, and evaluation and should avoid creating expectations that cannot realistically be fulfilled. Evaluation Assess whether participation processes improve policy quality, political agency, institutional trust, and democratic legitimacy. 8df08613-6b65-42c8-b9a9-8336bafc3a6d 4.6 Evaluators Assess participation quality, inclusiveness, transparency, efficiency, influence, and public outcomes. Researchers Generate evidence concerning the effects of AI on democratic participation and public trust. Evaluation should examine whether AI improves participation relative to clearly defined objectives rather than assuming that technological innovation alone represents progress. Continuous Improvement Refine participation methods through evidence, learning, experimentation, and adaptation while preserving democratic principles. 95a3cb5d-9f96-4e18-b1d0-a74b44fb7d80 4.7 Participation Practitioners Apply lessons learned to improve future participation processes. Public Innovators Develop, test, evaluate, and refine new methods that improve democratic participation. Improvement should be iterative and evidence-based. Successful innovations should be documented and shared, while unsuccessful approaches should also be analyzed so that lessons contribute to future practice rather than being lost. Political Agency Strengthen citizens' confidence that they can participate meaningfully and that their voices will be heard. 0e59d5b2-83bc-4bb6-84b3-69dd55f5d00b 4.8 Citizens Develop confidence that meaningful participation can influence public decisions. Public Institutions Demonstrate through their actions that citizen participation contributes to better governance. The OECD identifies political agency—the combination of confidence in one's ability to participate and confidence that one's voice will be heard—as one of the strongest predictors of public trust in government. AI should ultimately be judged by whether it strengthens rather than weakens that sense of agency. 2026-07-20 https://www.oecd.org/content/dam/oecd/en/publications/reports/2026/07/artificial-intelligence-and-the-future-of-citizen-participation_0608e00e/a1ee2e0a-en.pdf Owen Ambur Owen.Ambur@verizon.net