Governing with Artificial Intelligence

AI in civic participation and open government

Stakeholder participation refers to “all the ways in which citizens and stakeholders can be involved in the policy cycle and in service design and delivery” (OECD, 2017[140]). Civic participation, which focuses on engagement with the public, can yield benefits for governments and citizens alike. It not only builds trust in government, but it also reinforces democracy by providing meaningful opportunities for citizens to have a say in how government designs and implements policies and services (OECD, 2023[141]; 2024[142]; 2022[143]).With regard to AI policies and products, engagement early in the technology development cycle enriches the understanding of issues, helps align technological innovation with societal needs, and influences how well potential harms are mitigated (OECD, 2024[144]; UK Government Office for Science, 2023[145]).

Governments are using digital technologies to expand civic participation opportunities by enabling new forms and channels of interaction between citizens and public institutions. This contributes to the growing trend of Civic Tech: the use of digital technologies to reinforce democracy by enabling the public to be informed, participate in decision and policymaking, and increase governments’ responsiveness and accountability (OECD, 2024[146]). In particular, AI tools have relevant applications in civic participation and more broadly in open government policies and practices, including access to information, government communication and protection of the civic space.

Current state of play

Improving transparency and information access

AI’s ability to sort, filter and summarise vast amounts of information could lower barriers to disclosure, making it easier for governments to be transparent to the public (OECD, 2024[29]). For example, the US government is testing an AI-powered tool to review classified documents for future disclosure under the Freedom of Information Act (US FOIA, 2023[147]; MITRE, 2023[148]). AI can also help citizens understand complex governmental processes and create opportunities for broader engagement and scrutiny from citizens and civil society organisations (CSOs) (Araszkiewicz and Rodríguez-Doncel, 2019[149]).

Governments can offer AI-enabled tools to the public directly to help people navigate and make sense of extensive amounts of data and information, thereby providing citizens with tailor-made and accessible access to public information. For example, the District of Columbia (US) has released a beta version of the platform DC Compass,1 which uses AI to respond to users queries and generate maps based on the city’s open data. In Greece, the opencouncil.gr platform uses AI to automatically transcribe local council meetings and generate summaries, social media content and personalised neighbourhood updates via messaging apps, making local governance more accessible and understandable for citizens.2Another example from Greece is the platform DidaktorikaAI launched by the National Documentation Centre (EKT), improves the accessibility of academic and scientific knowledge for policymakers and the broader society through an AI-powered online library gathering more than 50 000 publications (National Documentation Center (EKT), 2025[150]). Furthermore, the European Parliament provides an AI tool to help citizens access its archive by asking questions in natural language (Box 5.35).

Box 5.35. Democratising access to the European Parliament’s archives

The European Parliament Archives have embraced generative AI to transform access to their extensive historical records. Developed by the Archives Unit, the innovative Archibot platform integrates LLMs with RAG techniques to enable users to interact directly with digitalised documents. Citizens, researchers and policymakers can now make natural language queries to retrieve information on historical events, key figures and institutional developments from an index of over 100 000 documents dating back to 1952. The tool employs RAG methods to identify relevant documents based on users’ queries and synthesizes the extracted information into concise, easily comprehensible responses. Accessible via the European Parliament’s archive website, the platform supports automated translation, thereby extending its reach to users in 55 languages despite the predominantly French content.

The project, which reduces document search and analysis time by approximately 80%, significantly improves the efficiency of historical research and data analysis. In addition to facilitating rapid access to legislative and policy information, Archibot supports the creation of comprehensive reports and analytical outputs, serving as a valuable resource for educators, researchers and policymakers. This digital transformation of archival access not only addresses longstanding challenges in navigating vast records collections but also enhances transparency and public engagement within the European Union.

Source: https://historicalarchives.europarl.europa.eu/en/sites/historicalarchive/home/cultural-heritage-collections/news/ai-dashboard.html.

Enabled in part by improved transparency and access to information, AI tools present opportunities to promote civic space (OECD, 2023[151]). Civic space is a cornerstone of functioning democracies. It consists of the set of legal, policy, institutional and practical conditions non-governmental actors need to access information, express themselves, associate, organise and participate in public life (OECD, 2022[152]). For example, in Israel, the platform Kol Zchut (All rights) — a comprehensive repository of residents’ rights developed by the homonymous CSO along with the country’s Ministry of Justice and National Digital Agency — has recently integrated an AI chatbot to allow users to formulate their queries in Hebrew, Arabic and Russian (OECD OPSI, 2019[153]). AI experts have identified empowered citizens, CSOs and social partners (e.g. trade unions) as one the most critical potential benefits of AI use, calling on governments to take more action to empower these actors (OECD, 2024[29]).

AI can also expand access to information by helping people with disabilities that may face barriers to their access to information. For example, citizens affected by visual or cognitive impairments can benefit from speech-to-text or image-to-speech conversion applied to government websites and portals, as well as from AI-powered accessible interfaces with adaptive texts, fonts, colours and contrasts (Welker, 2023[154]). The United Kingdom built its gov.uk website to be fully compatible with technologies that use AI to improve its accessibility. AI can also enhance access to information by overcoming the language divide, including on government websites through automatic translation of content, which has become mainstream. This can enhance the ability to engage in participatory processes by fostering dialogue among participants speaking different languages (European Parliament, 2022[155]). India’s Bhashini platform, discussed in Chapter 4, serves as another example of seeking to overcome language divides, including through leveraging participatory techniques.

Making participatory processes effective and inclusive

Governments can use AI to help conduct and facilitate public discussions and dialogues. Such dialogues often face a trade-off between breadth (the number and variety of participants) and depth of discussion and consensus-building (Landemore, 2022[156]). While the internet has enabled spaces for many-to-many communication, these spaces are not always designed to foster deliberation and consensus-building (Weyl, Tang and Plurality, 2022[157]). AI can provide for a welcoming and productive deliberative space by (Landemore, 2022[156]; Ovadya, 2023[158]):

• Extracting meaningful insights from the contribution of many participants, which could be provided under different formats, such as text and voice.

• Detecting of harmful content and supporting human moderators in countering hate speech.

• Facilitating fact-checking by providing data and sourced information in real time.

• Helping to perform live facilitation to favour multiple opinions and to create bridges towards consensus and support deliberation.

AI can also enhance the quality of dialogue. For instance, it can support organisers in drafting accessible and fact-based learning materials and simplify complex bureaucratic documents into easy-to-read language to enable informed debate. For instance:

• In the United States, the AI-powered tool MAPLE (Massachusetts Platform for Legislative Engagement) allows citizens to better understand the context and objectives of draft legal texts through AI-generated summaries submitting their inputs and comments.3

• In Spain, the participatory platform Decide Madrid, based on the open-source software Consul, experimented with an NLP system to assist citizens in aggregating and developing proposals (Arana-Catania et al., 2021[159]).

Research indicates that the use of AI-enabled deliberation tools could have a significant impact on decreasing polarisation (Tsai et al., 2024[160]; OECD, 2020[161]). The examples in Box 5.36 illustrate that such tools are not only beneficial for obtaining understandable insights on a specific topic, but can help bring together the results of the multiple deliberations. They can ensure consistency while preserving the richness of the debates (Landemore, 2022[156]), as well as strengthen connections between the participants and the broader society by making the content and results of discussions more accessible to the public.

Box 5.36. AI as facilitator of deliberative processes

Mass online deliberation with Polis

Polis is an AI-powered open-source software designed to allow large group of people to discuss, collaborate and reach consensus on complex issues. Participants share their opinions on a given question and express their agreement or disagreement on other participants’ statements. The algorithm maps participants’ responses and clusters opinions, facilitating the identification of positions that are supported by the majority of participants. Polis has been used to facilitate discussion and build consensus in Austria on environmental issues, in Uruguay on a national referendum, in New Zealand to facilitate the development of government policy, in the Philippines to generate municipal policy, in the United States to counteract polarisation at the municipal level, and in Germany to develop a political party’s platform.

In Chile, AI helps moderate deliberative processes

In 2020, the Senate of Chile partnered with the Stanford Center for Deliberative Democracy to involve citizens in the online deliberative process “LXS 400 – Chile Delibera” performed through the Stanford Online Deliberation Platform. The platform used an AI tool to keep time and distribute speaking rights equitably, allowing participants to form speaking queues and discuss in small groups with timed agendas. Over 500 citizens representative of the Chilean society discussed in small groups and deliberated in plenary sessions on health and pension reforms.

Source: (The Computational Democracy Project, 2024[162]; Fishkin and al, 2021[163]).

In addition to facilitating engagement, AI can help governments and citizens analyse and make sense of the inputs from participatory and deliberative processes. AI features — such as automation and processing of large amounts of data and text — can facilitate mass, multilingual online deliberation and dialogue. Governments usually face the challenge of analysing large quantities of qualitative inputs received in public consultations, which can be remedied with the help of AI. For example, the United Kingdom’s Incubator for Artificial Intelligence (i.AI) has developed the i.AI Consultation Analyser, which aims to expand the government’s capacity to analyse citizens’ inputs in legal public consultations. The tool is expected to reduce the cost of this activity by GBP 80 million (AI.GOV.UK, 2024[164]).4

AI sensemaking tools allow for the identification and evolution of recurrent topics, the automatic clustering of opinions and ideas, the detection of outliers, and the sentiment analysis of the perception of a given issue (Arana-Catania et al., 2021[159]; Berditchevskaia and Baeck, 2020[165]; Schneider and Sanders, 2023[166]). Sensemaking tools can help public servants understand and visualise citizens’ priorities and opinions on issues and support them in translating large volumes of inputs into actionable recommendations that are representative of the participants’ views. Citizens can also use these tools to understand others’ perspectives, inform and craft their own inputs, and better understand the results of participatory and deliberative processes. Examples include France’s national Grand Débat, or Great Debate (Box 5.37), as well as:

• The Cambridge City Council in the United Kingdom used the AI-powered sensemaking tool of the Go Vocal platform (formerly CitizenLab), which automatically clusters and prioritises citizen opinions to analyse their contributions to its Design Code, saving 50% of the estimated time for manual processing (Fillet, 2024[167]).5

• The French Economic, Social and Environmental Committee (CESE) organised a deliberative process (Convention Citoyenne) to inform legislation on the sensitive policy issue of end of life. The AI-powered tool Panoramic, developed by Make.org, allows citizens, journalists and CSOs to access and navigate the discussions and the results of the deliberative process via an interactive interface.6

Box 5.37. Improving participation in France’s Grand Débat National

In France, the Grand Débat National, launched in early 2019, exemplifies a pioneering effort to democratise public debate through digital innovation. In response to growing social discontent, the French government established this national consultation to gather citizen input on key issues — including employment, taxation and democratic reform — using a dedicated online platform. Artificial intelligence played a central role in managing and interpreting the vast array of contributions.

AI-driven techniques, such as NLP and clustering algorithms, were employed to systematically analyse millions of submissions. By identifying recurring themes and synthesising diverse viewpoints, these tools enabled policymakers to highlight consensus points and emerging trends. This automated analysis transformed raw citizen input into accessible insights, supporting more informed and transparent decision-making.

Source: https://granddebat.fr.

Finally, AI can facilitate innovation in civic participation activities by generating simulations and visualisations on various scenarios and alternatives, including in combination with other emerging immersive technologies like augmented or virtual reality (VR).7 Generative AI tools can help governments to engage with the public on complex projects concerning a physical space (e.g. urban mobility and redesign). These tools can simulate future and alternative scenarios, allowing citizens to visualise the final results and provide feedback for government consideration (OECD, 2023[2]). Generative AI can significantly reduce the costs of creating visual simulations by automating the design process and generating realistic renderings. For example, the City of Hamburg in Germany, used 3D and AI technologies in the CityScope platform, developed at MIT Media Lab, to engage relevant stakeholders in the decision-making process for 161 viable locations to home refugees (Campagnucci et al., 2025[168]). Similarly, Finland used the tool UrbanistAI to foster discussion on the pedestrianisation of streets (Box 5.38).

Box 5.38. Finland uses generative AI for participatory urban planning

In Finland, the City of Helsinki used generative AI to enhance participatory urban planning through UrbanistAI. This platform integrates AI-powered visualisation tools with co-design features that allow residents and businesses to collaboratively shape urban spaces. During the 2023 Summer Streets initiative, Helsinki used the platform in community workshops, where participants translated their ideas into AI-generated street designs, fostering more inclusive and engaging urban planning processes.

By enabling the real-time visualisation of community proposals, UrbanistAI helped bridge the gap between citizen input and final design decisions. The tool’s ability to align proposals with local policies also streamlined the planning process, ensuring regulatory compliance while maintaining creative flexibility. The success of the Helsinki case has inspired broader adoption of the platform, with cities including Tallinn, Pristina and Berlin integrating it into their participatory urban design initiatives.

Source: https://interoperable-europe.ec.europa.eu/collection/public-sector-tech-watch/reimagining-helsinki-participatory-urban-planning-generative-ai.

Supporting effective and tailored public communications

AI-powered analytics software can help public communicators monitor and analyse online debate, audience perceptions and attitudes on a given issue.8 AI can help them to improve the quality, quantity, accessibility and relevance of information governments deliver to different segments of the population. Moreover, AI tools can enhance the efficiency and effectiveness of public communicators by assisting in brainstorming, automating routine tasks such as media planning, and ensuring consistency in content across communication channels. These capabilities enable communicators to better engage with their audiences, tailor messages to diverse groups, and respond promptly to public concerns, thereby fostering a more informed and engaged citizenry. An example can be found in Box 5.39.

Box 5.39. Innovating public communications with AI in the United Kingdom

In the United Kingdom, the Government Communication Service (GCS) has been developing and testing GCS Assist, an AI-powered conversational tool designed to help government communicators improve the efficiency and effectiveness of their work. The tool integrates professional guidelines, communication frameworks and proprietary audience insights, ensuring that AI-generated outputs align with government standards. Because AI is most effective when used as a collaborative assistant, GCS Assist is tailored to support both experienced and less-experienced professionals by streamlining routine tasks, enhancing creativity and enabling more time for strategic work.

GCS Assist has undergone an iterative development process, beginning with an initial prototype piloted within the Cabinet Office in late 2023. The tool was designed to support government communicators by generating draft texts, plans and strategic ideas while ensuring data security within government systems. The alpha assessment in March 2024 rated the tool’s progress as Amber, acknowledging its potential while highlighting areas for further refinement. A phased rollout is now underway, starting with a limited group of users before expanding access more broadly.

Source: https://gcs.civilservice.gov.uk/wp-content/uploads/2024/04/2024-05-GCS-Innovating-with-Impact-Strategy-interactive.pdf.

The production and dissemination of information materials, as well as the assistance and response to citizens’ queries, is a fundamental yet resource intense function to enable effective participation. AI can also enhance public communications efforts by curating and tailoring communications materials. Generative AI can, for example, summarise and synthesise information and create content for different audiences in different formats (video, image, text, etc.). Moreover, GenAI can support governments in providing citizens with accessible and evidence-based information, allowing them to better understand the policy issues at stake as well as the role of the government in regulating them. Examples from the Netherlands and Brazil in Box 5.40 illustrate how this can be done in government.

Box 5.40. Practices of AI for information curation in citizen participation

Helping reporting oral deliberations in the Netherlands

The Netherlands House of Representatives implemented the Speech2Write system, which allows its Parliamentary Reporting Office (PRO) to convert voice to text, as well as translate voice into written reports and produce an "almost-ready" report for human finalisation. The system is based on a parliamentary language model that was developed from scratch based on 2000+ hours of audio of proceedings and text files.

Making complex information more accessible in Brazil

The Brazilian Chamber of Deputies is using AI to improve the support given to parliamentarians and to build a more public-friendly legislative institution. Its Ulysses AI Suite uses an NLP system to classify and help citizens navigate the vast amount of information produced by the Chamber, including laws, speeches, data and research documents. Ulysses also helps parliamentarians draft legislative texts and find similar bills already adopted or rejected.

Source: https://www.ipu.org/innovation-tracker/story/artificial-intelligence-innovation-in-parliaments, (Arana-Catania et al., 2021[159]), https://www.camara.leg.br/noticias/548730-camara-lanca-ulysses-robo-digital-que-articula-dados-legislativos.

Virtual assistants such as chatbots can support communications efforts by helping citizens to navigate government information, by responding to their queries in plain language, and by sending personalised notifications and content based on their interests (Androutsopoulou et al., 2019[169]; van Noordt and Misuraca, 2019[170]; Cortés-Cediel, 2023[171]). In some cases, chatbots have been used as a channel for civic participation (see examples in Box 5.41).

Box 5.41. Chatbots supporting participation in Spain and Colombia

In 2021, the city of Madrid (Spain) embedded Clara, an AI-based virtual assistant into its existing participatory platform, Decide Madrid, based on the open-source software Consul. Clara responds to users’ questions on the functioning of the platform and of the participatory processes of the city.

In 2022, the city of Bogota (Colombia) deployed Chatico. This virtual assistant supports the interaction between citizens and the local administration by improving the experience of procedures and services and enabling citizen participation. Chatico is available both on the website and via WhatsApp to enhance its accessibility.

Source: https://www.madrid.es/portales/munimadrid/es/Inicio/Actualidad/Noticias/Nace-Clara-el-asistente-virtual-de-Decide-Madrid, https://oecd-opsi.org/innovations/chatico-virtual-agent, https://gobiernoabiertobogota.gov.co.

Strengthening electoral processes

Policy discussions on AI’s potential in the electoral process have focused on protecting electoral processes against the risks of AI, namely by empowering society and building resilience, protecting democratic principles, and promoting information integrity. Yet there are also relevant opportunities for AI to strengthen electoral processes by pre-empting fraud and disenfranchisement, and lowering barriers to entry for underfunded candidates (Eisen et al., 2023[172]; OECD, 2024[29]).

AI tools can be used by electoral management bodies (EBM) throughout the electoral period (Juneja, 2024[173]). Before elections, AI tools can be adopted to support voter list management, voter registration, resource allocation planning, forecasting election costs or targeted advertising. For example, in the United States, the Electronic Registration Information Center (ERIC) (2024[174]) uses ML to analyse voter registration and motor vehicle records. This helps states identify duplicate registrations, deceased voters and eligible but unregistered individuals, which improves voter roll accuracy. During elections, AI tools can contribute to campaign and media monitoring, voter verification, polling place monitoring and ballot counting. Finally, AI can support authorities in post-electoral audits by detecting incidents or frauds (Juneja, 2024[173]).

Managing risks and challenges

AI offers significant opportunities to empower citizens and civil society and reshape relationships between governments and their people for the better. However, potential risks persist, which can have direct implications on democratic spaces. While technological advancements may progressively address these issues, policymakers should consider them when adopting AI tools in this area.

Associated risks

• Inadequate or skewed data in AI systems

• Overreliance on AI

• Lack of transparency and explainability

• Resistance: Lack of public understanding about how government uses AI

• Misuse or questionable use of AI, resulting in surveillance and privacy concerns

• Exacerbating social exclusion and digital divides

If AI systems rely upon inadequate or skewed data, it could lead to inaccurate or adverse outcomes for some individuals or groups. With regard to civic participation and open government, this could result in inaccurate or imprecise consideration of citizen input into AI-enabled participation processes.

Moreover, deliberative processes are meant not only to harness collective intelligence and enable societal dialogue on specific policy issues, but also to enhance the knowledge of participants as well as their mutual understanding and empathy (OECD, 2021[175]). Overreliance on AI tools to improve the efficiency of deliberative processes might overlook other fundamental aspects of deliberation.

AI’s lack of transparency and explainability, especially in complex systems, can decrease trust in AI tools, and when these are used in democratic spaces, it can in turn affect trust in the participatory process and its outcomes. To address this challenge, some governments have launched algorithm registers that disclose detailed and technical information about algorithms, such as, their purpose, design, data inputs, decision-making processes, potential biases, among others (for details and examples, see Chapter 4, section on “Establishing guardrails to guide strategic and responsible AI”).

In addition, there are still many unknowns about user perceptions regarding AI in participatory processes and the influence it may have on their contributions and their willingness to participate. Experiments are being conducted to assess the reaction of participants to the introduction of AI tools as intermediaries in participatory and deliberative processes (Hadfi et al., 2021[176]; Kim et al., 2021[177]). The use of AI tools to support participants in formulating inputs and contributions should be carefully handled to avoid that these tools compromise the creativity in language and reflection of participants, focusing rather on efficiency and agreement.

Some governments have misused digital technologies for surveillance or even silence populations and digital opposition, thus damaging the civic space (OECD, 2022[152]). The use of AI systems for surveillance purposes, content censorship and improper forms of predictive policing threaten the free exercise of civil freedoms, such as the right to peaceful assembly and freedom of expression. These improper uses risk creating general mistrust among citizens towards the adoption and deployment of AI systems in government functions and public services. The OECD (2022[152]) found that most national AI strategies fail to discuss the impact of AI on the ability to freely exercise rights, although around half propose concrete oversight and redress mechanisms (OECD, 2022[152]).

Finally, many languages are insufficiently represented in AI systems, which are mainly trained in English, Spanish, and Mandarin (Peixoto, Canuto and Jordan, 2024[178]) (see Chapter 1, section on “Exacerbating digital divides” for a detailed discussion of the issue). In the context of citizen participation, this means that inputs submitted in other languages might not be processed and valued in the same way, creating new democratic imbalances (Romberg and Escher, 2024[179]). For example, to address the language divide and to preserve the Icelandic language, the government of Iceland partnered with OpenAI to train the LLM GPT-4 in Icelandic (Government of Iceland, 2023[180]). Similarly, the University of Turku (Finland) partnered in 2023 with the company SiloAI to build the Poro suit of systems, a family of multilingual open-source LLMs for all European official languages (University of Turku, 2023[181]). Similar efforts exist in for other languages, including Indigenous and endangered ones (OECD, 2023[182]).

Implementation challenges

• Skills gaps

• High costs of AI adoption and scaling

The adoption of AI tools by civil servants in charge of participatory and deliberative processes, as well as by citizen participation practitioners, requires adequate skills. This skillset spans from the identification of relevant uses of AI tools to support participatory and deliberative practices (OECD, 2025[183]), to the selection and purchase or development of AI solutions that meet the defined objectives, to the use of AI tools in the context of participatory and deliberative processes, to the evaluation of the quality of their outputs.

The adoption, either through in-house development or through public procurement, and the use of AI tools to support citizen participation requires a flexible approach including pilot initiatives, testing, and iterative adjustments. Such approach entails significant costs and require an elevated high tolerance to risk. Challenges related to costs and scaling of AI tools are particularly relevant for citizen participation, ss participatory processes are often organised and implemented by local governments (OECD, 2022[184]).

Untapped potential and way forward

Given the novelty of the use of AI in civil participation and enablement, it is difficult to assess evidence of impact and general lessons learned beyond individual case studies. The use cases showcased in this chapter suggest that AI has the potential to improve the efficiency, reduce the costs and enlarge the scope of participation, ultimately lowering barriers for citizens and CSOs to voice their opinions and provide feedback at various stages of policymaking. The use of AI tools to process citizens’ inputs, automate tasks, support the creation of information and communication materials, and provide citizens with 24/7 virtual assistance could significantly reduce some of the costs associated with the design and implementation of relevant processes. These savings, coupled with the growing availability of open-source tools, could enable citizen participation opportunities in contexts with limited resources, such as in smaller cities and less developed regions.

Digital technologies have enabled asynchronous participation and significantly reduced the burden of geographical barriers, allowing for broader collaboration (OECD, 2022[184]). AI has the potential to lower other barriers, namely the language divide, allowing higher levels of citizen participation. Moreover, virtual assistants and chatbots can further improve the accessibility of participatory and deliberative processes by helping participants navigate complex information. Finally, AI tools for moderation, facilitation and sense-making could become essential to addressing the challenge of scale in deliberative processes (Landemore, 2022[156]).

Early efforts could lead to longer-term conceptions of transparency and responsiveness. Digital technologies have long affected how individuals engage with notions such as their so-called “right to know,” and voice opinions about complex operations and their outcomes (Margetts, 2011[185]; Song and Lee, 2015[186]). AI technologies also offer novel ways of enhancing service design, delivery and transparency. Individuals may begin to find current operational systems antiquated and difficult to use, and therefore they may voice their opinions about how services should integrate technologies that can enhance access to information (Carrasco et al., 2019[187]). This could help accelerate institutional reforms that aim to improve transparency and accessibility. These may be driven not only by the adoption of AI technologies but also by their capacity to shift public perceptions and expectations of institutional capacity.

Moving forward, governments should explore AI’s potential to enhance existing civic participation processes by expanding government capacities of analysing citizens’ contributions, enabling mass-deliberation and reducing barriers to participation. Governments should carefully consider the challenges associated with the use of AI in civic participation, namely issues related to increased technological intermediation of interactions between citizens and government, and its impact on public discourse and trust. The adoption and deployment of AI tools for civic participation should follow clear strategies that include transparency and accountability as fundamental principles. Citizens and civil society should be actively involved in the elaboration of such strategies to ensure alignment with democratic principles and enhance trust and acceptability of the process (see Chapter 4, section on “Engagement to shape strategic and responsible AI”).

Government services and justice functions

Government services and justice functions are essential to meeting citizens’ needs and upholding the rule of law. This includes the design and delivery of public services that are accessible, efficient and responsive, as well as effective law enforcement and disaster risk management to ensure safety and resilience. Justice administration and access to justice are equally critical, providing fair and timely legal processes that protect rights and promote public trust. These areas have the most impact with the public. Governments are using AI in a variety of ways to ensure that they are efficient and effective, and that citizens and residents are safe and treated fairly.

Notes

← 1.  https://opendata.dc.gov/pages/compass.

← 2.  https://opencouncil.gr.

← 3.  https://www.mapletestimony.org.

← 4.  https://ai.gov.uk/consultations.

← 5.  https://www.govocal.com.

← 6.  https://panoramic.make.org/customer/cese/event/convention-citoyenne-sur-la-fin-de-vie.

← 7.  See https://oe.cd/immersive-brief for an OECD briefing document on immersive technologies.

← 8.  The key opportunities for public communication identified in this paragraph are the results of interviews conducted by the OECD in 2023-2024 with members of the OECD Public Communication Network.