OECD Regulatory Policy Outlook 2025

The scale and pace of innovation are fundamentally changing the way that societies and economies function. Emerging technologies such synthetic biology, artificial intelligence (AI), advanced materials, neurotechnologies and quantum technologies can contribute to unprecedented gains in health, energy, climate, food systems and biodiversity (OECD, 2024[1]). However, these innovations can also bring risks for people, the economy, the environment and democracy. Rules and regulations, whether developed by government or written in collaboration with or entirely by industry, provide avenues for managing these risks while supporting innovation. By acting as a “gatekeeper of the market” (Evans, 20 January 2021[2]), well-designed rules help harness innovation to promote economic, social and environmental goals.

But regulation is not always in place, effective in ensuring the necessary protections or perceived as appropriate for driving a positive impact. In 2024, data from 30 countries show that over a third of citizens find it unlikely that their national government would appropriately regulate new technologies and help businesses and citizens use them responsibly1 (OECD, 2024[3]). Without timely and informed regulatory action, gaps in protections and market functioning can emerge. This can lead to increased risks; hindering the responsible adoption of new technologies; and leaving both markets and individuals vulnerable to misuse, exploitation or inefficiencies.

Digital technologies are particularly pertinent to current regulatory discussions, as technologies such as AI, the Internet of Things (IoT) and quantum technology rapidly evolve and transform our everyday practices and future potential while creating or exacerbating potential harms. Well-cited examples of risks linked to the use of digital technologies include facial recognition and spyware as tools for mass surveillance (Ryan-Mosley, 2022[4]); cyber-attacks and cyber-crime undermining citizen privacy and safety; social media platforms as a vector for misinformation (Matasick, Alfonsi and Bellantoni, 2020[5]; OECD, 2024[6]); and biased algorithms leading to discriminatory hiring practices, for example, based on race and gender (Chen, 2023[7]).

Regulation plays an important role in ensuring that governance systems are sufficient to align the development and application of digital technologies with positive societal outcomes. While industry itself should aspire to this, the right incentives do not always exist. Industry-led or co-led approaches can enable more agile responses to technological change and lower information asymmetries, but their practical implementation has, at times, left the public ill-protected including by prioritising innovation over other regulatory objectives (OECD, 2024[6]). For example, X (formerly known as Twitter) withdrew from its voluntary participation in the 2018 European Union Codes of Practice on Disinformation in May 2023 (OECD, 2024[6]). Governments thus play a critical role in guiding and enforcing responsible digital transformations through their regulatory powers to make risk-informed policies.

Regulating for the future requires governments to understand and plan responses to the current, emerging and future challenges – the most salient currently being the twin challenge of green (see Chapter 3) and digital transitions. This chapter provides insights on how governments can regulate for the future of digital transformation, from which lessons can be applied to other, future transitions. The goal is to demonstrate the benefits of regulation as a valuable tool to unleash the positive impact of innovation responsibly, supporting and building upon broader OECD initiatives (Box 4.1).

The good news is that decades of regulatory reforms have resulted in a system of processes, tools and institutions to help maximise the benefits of regulating while minimising costs to citizens, businesses, society and the environment (OECD, 2021[8]). The key question becomes how to adapt these three elements to ensure regulatory governance is fit to tackling emerging issues? This chapter examines the complex challenges facing regulators in the digital age and presents concrete steps governments can and are starting to take to develop rules and frameworks that are fit for the future:

• adapt processes for responsive regulation

• harness novel tools to improve regulations

• shape future-ready regulatory institutions.

Digital transformation is one of the most pressing complex challenges for policymaking. On the one hand, many OECD governments are looking at their existing systems of better regulation and concluding that several elements within them still function effectively in the digital age. This is supported by OECD research on regulatory approaches to AI demonstrating how existing systems are working to identify and react to the risks brought about by AI systems (Box 4.2). On the other hand, discussions across international fora have begun to identify gaps in governance and regulatory approaches when faced with the risks of digital transformation and challenges to support emerging opportunities (see Box 4.1).

Two major challenges governments face when regulating in the digital age are the pace of technological development and the transboundary nature of digital technologies. Both challenges set the foundation for the regulatory adjustments and improvements presented in the rest of this chapter to effectively regulate for the future.

Effective regulation requires a shared understanding of the policy problem, not only among government entities but by all stakeholders. The faster technology develops, the harder it is to build this shared understanding. Information asymmetries on the opportunities and risks of technological development, coupled with uncertainty on its trajectory, can quickly accumulate, making it difficult for policymakers to take informed governance decisions on the need for regulatory interventions, what to regulate and how to do it effectively.

Due to their rapid development, the impact of many digital technologies may not be fully understood until years after their creation. Technology’s intangible nature and the increasing convergence (e.g. the combination of digital technologies with physical ones) result in complex relationships whose impacts are difficult to predict or measure. For example, the combination of AI and the IoT technologies could potentially introduce new cybersecurity threats alongside their benefits, requiring updated regulatory frameworks to better ensure the security of these technologies. Without a thorough understanding of the challenges and gaps created by technological innovation, policymakers are ill-equipped to design governance approaches that target a specific need.

Further, continuously evolving value chains and business models complicate accountability over digital technologies and their ultimate outcomes. Governments can struggle to understand how to best regulate a technology to incentivise appropriate accountability structures. For example, should policymakers focus interventions on digital technology development, its application or its ultimate impact? Each has a distinct nature that inevitably influences policy and regulatory responses.

Rapid technological development can also challenge the adequacy of current regulatory approaches. Many governments are questioning whether the existing rules and approaches are sufficient for the challenges posed by digital technologies – from sectoral and/or horizontal perspectives. Regulation is traditionally designed issue-by-issue, sector-by-sector or technology-by-technology and often with a “set and forget” approach (OECD, 2021[8]). The rapid pace of technological progress challenges this model, requiring shorter time frames and moving away from ex ante design and ex durante delivery as a series of discreet steps or tools but rather mutually complementary parts of the policy cycle meant to inform the adaptation of regulatory (or alternative) approaches (OECD, 2024[6]; 2021[8]). This is further aggravated by the Collingridge Dilemma, in which rules on technologies are easier to accept when the technology is in early-stage development, though impacts are hard to know, compared to later-stage development when rules on technologies are harder to accept but there is greater evidence of their impacts (Tõnurist and Hanson, 2020[21]).

Technologies and their applications cut across sectors and government institutions, creating a landscape where the governance of digital innovations often falls within the purview of multiple policymakers and regulators. This phenomenon results in a complex web of responses and responsibilities that needs to be managed. Overlapping jurisdictions and regulatory gaps can lead to inefficiencies, inconsistencies and missed opportunities in governance. Addressing policy problems now, more than ever, demands a joined-up and collaborative approach between regulators to identify the existing mandates and gaps in institutional design for technology governance. The development of international standards and enhanced information sharing between countries is seen as beneficial to effectively mitigate digital threats, while existing frameworks can also be continuously adapted to new technologies to remain robust.

Technologies are also challenging traditional notions of legal liabilities (OECD, 2024[6]). This includes questions about jurisdiction, mandates of regulators to enforce rules, and blurring the boundaries between consumers and producers (OECD, 2024[6]). This shift not only challenges the applicability of traditional legal concepts but also makes it difficult to craft effective policies that account for these new dynamics. Moreover, the cross-border nature of digital technologies further complicates jurisdictional authority, as actions in one country can have significant impacts in another, requiring a more nuanced and collaborative approach to regulation that transcends national boundaries. In this context, international regulatory co-operation becomes essential to ensure consistency, prevent regulatory fragmentation and create a more coherent framework for addressing issues that span multiple jurisdictions.

Finally, there is a challenge in building institutional capacity to think long term about how emerging technologies may affect societies, markets and government actions (OECD, 2024[6]). This requires much broader use of anticipatory regulatory approaches, including increasing the capacity of oversight and advisory bodies to anticipate and implement strategic foresight around the policy cycle. This also requires that governments give agency to public servants and an authorising environment that validates anticipatory innovations (Tõnurist and Hanson, 2020[21]), as well as developing the necessary skills needed to foster evidence-based policymaking for digital technologies (OECD, 2024[6]; 2021[8]). OECD reviews focusing on anticipatory innovation governance have highlighted how Finland, Ireland, Slovenia and others are attempting to build institutional capacity for innovation (OECD, 2022[22]; 2021[23]; 2021[24]).

To manage both the pace and transboundary nature of digital technologies, governments need to be flexible and agile in their regulatory processes. However, regulators have often adopted a regulate-and-forget mindset, developing policy solutions to policy problems but then failing to monitor, evaluate or update them over time. Successive OECD Regulatory Policy Outlooks have noted that OECD Members continue to lag when it comes to ex post evaluations, relatively to the other regulatory management tools (OECD, 2021[8]; 2018[25]; 2015[26]). In the context of the digital transformation, such a static approach can leave regulatory regimes outdated, unfit and overly burdensome to tackle modern challenges.

Rather, what is needed is an adapt-and-learn process that enables government to continuously learn from and improve its regulatory approaches and systems (OECD, 2021[11]; 2024[1]). This approach allows governments to remain responsive to the fast-changing nature of digital technologies, ensuring regulations remain relevant and effective. Key elements to achieve such a responsive regulatory system are the employment of anticipatory governance and an iterative policy cycle.

To better inform decision making and equip institutions to govern digital technologies, governments need to adopt an anticipatory approach. This approach seeks to address technology as it emerges and evolves to increase the power of governance to stimulate innovation while better aligning innovation and regulation trajectories with societal goals.

The OECD (2024[1]) Framework for Anticipatory Governance of Emerging Technologies elaborates on how to move from managing technological risks to “getting ahead” of technological developments (Guston, 2013[27]). Doing so requires governments to consider five interconnected elements to apply in specific technology contexts:

By incorporating these elements of anticipatory governance, governments are better equipped to proactively address emerging challenges; adapt to technological advancements; and create more resilient, forward-looking policies that can navigate future uncertainties. Two particular processes – strategic intelligence and stakeholder engagement – are outlined below.

With growing complexity, and uncertainty, it is important that governments build a knowledge base on the potential evolution of digital technologies and their impacts. A lack of foresight can leave governments flat-footed when crises, such as COVID-19, emerge or when new technologies disrupt everyday processes. Famously, Uber disrupted the regulatory regimes governing taxis in cities around the world – the emergence of which regulators had not adequately foreseen or planned for. Strategic intelligence approaches prepare both policymakers and regulators to adapt regulatory systems so that they are resilient and prepared for potential change.

Strategic intelligence methods include horizon scanning, technology or strategic forecasting, foresight, technology assessment, and emerging risk assessments (OECD, 2024[1]). Box 4.3 provides guidance on the general use of these processes. From an agile regulatory policy perspective, governments commonly employ two connected approaches: horizon scanning and strategic foresight.

Horizon scanning is the detection and analysis of weak signals of technological developments. Horizon scanning is the foundation of any strategic intelligence process. It helps pinpoint areas of further interest and understand the key drivers of technological change. It can involve desk research, expert surveys and a review of existing futures literature. It can also involve megatrends analysis, which explores and reviews large-scale changes building in the present at the intersection of multiple policy domains, with complex and multidimensional impacts in the future.

Strategic foresight draws on multiple data sources to expand on potential, alternative future scenarios and their implications for policy. Policymakers use it to draw greater linkages between advancements in technology and their implications for governance. This is particularly useful for government scenario planning – developing multiple stories or images of how the future could look and using this to support informed decision making in the present.

Many OECD governments are adopting horizon scanning and strategic foresight to improve their regulatory policy. Instead of relying on occasional foresight efforts, countries are now establishing dedicated groups or advisory bodies focused on this task to help inform their decision making. For example, the United Kingdom has dedicated resources for horizon scanning to support rule-making for quantum technologies (Box 4.4).

In addition, other OECD Members have created or appointed a body dedicated to conducting regulatory foresight:

• Portugal has set up a Competence Centre for Planning, Policies and Foresight of Public Administration covering the full regulatory cycle.

• Korea’s Regulatory Reform Committee, together with ministries, has defined a “Pre-emptive Regulatory Innovation Roadmap” to proactively identify and address regulatory issues related to emerging technologies.2

• The European Commission has taken steps to integrate strategic foresight into EU policymaking, including as part of the Better Regulation Agenda. In addition, the Strategic Foresight Network ensures long-term policy co-ordination between all Directorates-General, and the European Commission is co-operating on foresight with other EU institutions through the European Strategy and Policy Analysis System.

• Germany’s Federal Chancellery has developed foresight capacity in recent years, notably in connection with the effects of technological innovations, such as cyberattacks as an unconventional military means or the role of social media manipulation in election campaigns.

The next step for OECD Members is to make foresight and scanning a regular part of policymaking, ensuring it has a broad impact and is not limited to small expert groups or isolated projects (OECD, 2020[30]). This includes engaging a range of stakeholders to help validate findings and identify regulatory implications from the foresight results and incorporating findings into the design, delivery and review of new and existing rules (OECD, 2021[31]).

Stakeholder engagement is a critical process that helps governments keep abreast of the developments and effects of technological changes (2022[22]). However, at least two challenges currently impede governments from effectively using stakeholder engagement in regulating digital technologies. First, stakeholder engagement often happens late in the regulatory process, after key decisions have already been taken, missing the chance for ongoing input from stakeholders throughout the policy cycle. Second, there is a growing difficulty in ensuring that governments collect the information required to effectively govern. Regulators face an information asymmetry regarding the current and future capabilities of technologies, making it difficult to develop fit-for-purpose regulation over the long term. Firms, civil society organisations and citizens are on the frontline of these changes and can offer valuable inputs on the feasibility of solutions (OECD, 2021[8]). Engaging stakeholders can enrich the understanding of issues by contributing missing knowledge, opening problem framings, illuminating key values at stake and anticipating barriers to effective implementation (OECD, 2024[1]).

Stakeholder engagement is thus a critical step to ensure well-designed regulations that anticipate technological advancements from diverse perspectives. To bridge the information gaps, governments need a partnership model, working directly with engineers and developers to understand technology trends and functionality, co-designing technology strategies and agendas, encouraging communication through interdisciplinary and transdisciplinary processes such as research and development, and establishing collaborative platforms to nurture emerging technologies and strengthen the link between innovation processes and their societal impacts (OECD, 2024[1]).

Forthcoming OECD research on regulatory approaches to AI identifies ways in which governments are already considering how to address the first challenge.3 Australia established an expert advisory body to consult with industry to develop AI Safety Standards and new guardrails to support safe and responsible AI. Canada’s proposed AI and Data Act4 would mandate ongoing dialogue with industry experts, academic researchers and international bodies in the implementation of its policy via self-regulatory and standards-based practices. The European Union’s AI Act creates the European Artificial Intelligence Board,5 a permanent co‑ordination platform and advisory body to the European Commission composed of experts and stakeholders. To anticipate the AI Act’s entry into force, the European Union also established the AI Pact that seeks industry to voluntarily commit to the AI Act and start implementing its requirements ahead of the legal deadline. The pact is implemented via a network of participants to share best practices and information (European Commission, 2024[32]). Israel’s AI policy includes forums for public participation and regulatory discussion. The United Kingdom’s pro-innovation approach to AI regulation highlights the need for central monitoring and feedback loops to ensure the regime is effective and adaptive and commits to implementing continuous feedback mechanisms. The United States’ Executive Order on AI instructs various agencies to solicit input from stakeholders for studies, pilot programmes and regulatory recommendations. Complementary information about the implementation of AI policies around the world can be found in OECD (2024[33]).

Beyond national governments, international organisations are also playing a key role in facilitating stakeholder engagement on AI. In the monitoring of the OECD (2019[9]) Recommendation of the Council on Artificial Intelligence, the OECD integrated input from a broad range of stakeholders, in particular through its AI Group of Experts (OECD, 2024[33]). The OECD will similarly engage stakeholders in the monitoring of applications of the Hiroshima Process International Code of Conduct for Organisations Developing Advanced AI Systems (OECD, 2024[34]). The Council of Europe’s (2024[35]) Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law, adopted in May 2024, and the United Nations Global Digital Compact (United Nations, n.d.[36]), adopted in September 2024, both use extensive multi-stakeholder engagement efforts.

It is not enough for governments to anticipate regulatory and governance needs. They need to adapt the processes throughout the policy cycle – regulatory design, delivery and review – to ensure that the regulatory system can effectively manage responsible digital technologies. In practice, governments have struggled to adapt regulations: by the time regulators identify problematic areas, the technologies have already evolved, making it challenging to impose meaningful control. As advancements in digital technologies have progressed, regulators have often been caught reacting to issues after they arise, rather than proactively shaping policies to guide technological development. AI has been a particular area of concern, as it poses difficult issues on privacy, safety and security, among others.

Controlling these issues is made even more difficult by the complex network of actors involved in the development and use of digital infrastructure, platforms and applications. From tech companies and software developers to platform users and content creators, the ecosystem of digital technologies involves a vast array of stakeholders, each with different interests and responsibilities. For example, digital platforms such as social media amplify the reach of harmful digital content, such as hate speech, illegal activities and disinformation often via their own algorithms. This complexity makes it hard for governments to pinpoint where regulation should be directed and how to enforce compliance effectively. By establishing iterative processes within the policy cycle, government can better understand the complexity of digital technologies, identify the regulatory policy gaps and design more effective regulatory systems for responsible digital development.

Effective regulatory design for digital technologies acknowledges and addresses the inherent differences both across and within these various technologies. Digital technologies evolve at different rates and can have drastically different impacts depending on how they are applied. For example, quantum technologies are more advanced with regards to sensing and timing capabilities, but more nascent when it comes to computing abilities. This poses different risks and benefits depending on this context. As a result, regulatory frameworks must be flexible enough to accommodate this diversity, ensuring that they are neither overly restrictive nor too lenient, and capable of evolving alongside technological innovation.

There is no one-size-fits-all solution when it comes to regulatory design and no comprehensive list of regulatory approaches, because the optimal design choice depends on the specific digital technology and the context in which regulation is being considered. However, certain regulatory strategies have shown promise in bridging existing gaps while maintaining enough flexibility to support further innovation.

One common strategy many governments employ is the risk-based regulatory approach. This approach focuses on differentiating the intensity of regulation based on the deemed level of risk of an innovation. The European Union’s AI Act is a prime example of this approach, as it classifies AI applications based on their risk levels and imposes different degrees of regulation accordingly. High-risk AI applications can face restrictions or even outright bans if deemed too dangerous, ensuring that while innovation proceeds, the risks to society are minimised.

Another regulatory strategy often considered is an outcomes-based approach. This type of approach focuses on achieving or preventing specific outcomes rather than prescribing detailed processes or technologies. This design prioritises flexibility, allowing industries to adopt whichever methods best achieve the required regulatory goal. Outcomes-based regulation is most commonly applied to innovations where targets on performance relating to costs, reliability, safety, etc. can easily be set and monitored. However, further research is needed to provide evidence on their effectiveness for digital technologies.

A key feature of several of these approaches is an emphasis on technology neutrality, which aims to abstract from regulations that encourage a particular type of technology, treating technologies similarly to the extent that they have the same effect. Tech neutrality can make regulations more resilient to technological change and more adaptable to evolving regulatory environments, making it an important principle to consider in regulatory design. However, as with any regulatory approach, there may be contexts in which deviations need to be considered to address challenges posed by specific technologies.

Regulatory impact assessments (RIAs) should not only consider the immediate effects of regulation, but also the impacts that regulation may pose to future developments. In the case of digital technologies, regulation should manage the risks, while ideally not stifling future, positive innovations. Updating the methodology and guidance for ex ante impact assessment presents a unique opportunity to embed innovation-orientated thinking into rule-making. By enhancing this assessment process, policymakers can be encouraged to consider factors such as new innovative solutions and technological change, and to adopt best practices for effective implementation.

However, most OECD Members have yet to adapt their RIAs to include elements of innovation-related impacts. As shown in Figure 4.1, about a third of OECD Members reported that their RIA system addresses innovation-related challenges. In these cases, adjustments were typically made by updating the RIA process, such as revising templates or providing explicit guidance on assessing the impacts of regulation on innovation via, for example, experimentation and outcome-based regulation. Box 4.5 details how countries are adapting RIA to anticipate innovation.

A key element of an effective regulatory policy cycle is to engage in frequent monitoring – creating feedback loops to continuously assess and refine regulations. This helps governments adjust regulations in response to new information and ensure that policies remain relevant and effective over time, as showcased by New Zealand’s regulatory stewardship in Box 4.6.

Reviewing the existing regulatory stock is essential to ensure that current frameworks remain relevant and effective in the face of rapidly evolving digital technologies. As digital technologies reshape industries and introduce new risks, it is critical to assess which regulations are working, which have become outdated and where there are gaps that need to be addressed. This process not only strengthens the current regulatory system by refining or eliminating ineffective measures but also provides valuable insights for designing new governance approaches.

One element of ex post evaluation is assessing whether regulations have been effective in achieving their goals with regards to digital technologies, such as increasing the technology’s transparency and accountability over outcomes or limiting negative impacts. The other element is evaluating whether regulations can still support the innovation process. As demonstrated by Figure 4.1, reforms to integrate innovation-related challenges into ex post evaluation similarly adopt this trend, with fewer countries adopting such reforms. Nonetheless, some OECD Members are focusing on evolving their ex post evaluation systems to better tackle the challenges posed by innovation. Box 4.7 demonstrates efforts in Canada and the United Kingdom to conduct reviews aimed at promoting innovation via better regulation.

Novel tools, often powered by digital technologies themselves, are transforming how governments can inform and manage regulatory systems for the future. Advanced data analytics, foresight and scenario planning, and regulatory experimentation generate new insights on which governments can take informed regulatory decisions. Policymakers across the OECD are increasingly integrating these tools to fill evidence gaps and design rules that are better prepared for the future. This ultimately can improve the effectiveness of regulation; reduce compliance burdens; and create more informed, responsive decision making.

While new digital technologies can challenge regulatory processes and existing regulatory regimes, they also offer opportunities to enhance how rules are made and delivered. Technology can be leveraged to significantly improve the quantity and quality of data that governments have at their disposal throughout the policy cycle. Digital tools can make data available from new sources, sometimes in real time.

For example, AI applications can greatly bolster how, and how much, data can be analysed. These new insights can be game changers for better and faster decisions based on comprehensive knowledge of the regulated environment, provided that the right infrastructure is in place (OECD, 2020[45]). AI can be used in the design of regulations via anticipatory analysis of future scenarios and risks, assessing and experimenting with policy options, improving drafting and legislative fragmentation, and supporting the use of evidence-based policy tools. In regulatory delivery, it can help regulators model risks to improve inspections, detect non-compliance, monitor the evolution of risks and use data-driven methods that improve the way they deliver their mandates as world-class institutions. These are discussed more fully in forthcoming OECD papers on AI in regulatory design and delivery.

Using technological solutions early in the policy cycle can equip policymakers with better data to choose interventions that are more likely to have the desired impact. Large data sets – unfeasible to compile or analyse without the help of technology – offer a more comprehensive overview of the policy landscape and provide evidence on the potential impacts of different policy options (Box 4.8).

Chatbots, including AI-powered ones, can facilitate public consultations by interacting with many stakeholders simultaneously, gathering feedback efficiently and synthesising data. Chatbots can provide instant responses to stakeholder queries, guide them through the consultation process and compile their feedback, making it more accessible and lowering burdens for participatory policy design. In Estonia, an AI-powered virtual assistant, Bürokratt, was created as a single channel for public services and information. However, the effectiveness of AI systems, including chatbots, must be evaluated over time to avoid unforeseen failures, fine-tune them to maximise their benefits and understand their training data to mitigate potential biases. Chapter 2 further discusses using technology to enhance consultation practices.

Digital technologies can also be used to augment RIAs. In Germany, the Service Centre for Better Regulation in the Federal Statistical Office has proposed AI tools to support the estimation of compliance costs. This approach uses AI to scrape legal texts and make predictions on which new legal text changes compliance costs using high/low estimates. If the costs are low, they use AI to derive the compliance costs but if the costs are high, this is still conducted manually. However, there are challenges relating to sufficient technical equipment, the structure of data scraped, the understandability of German legal texts, data quality, explainability of variables used in the model and matching across data sources (Walprecht and Lewerenz, 2024[47]).

Data that were previously inaccessible or only usable at significant administrative cost can be harnessed through technology to enable more effective monitoring of rules in practice (OECD, 2020[45]). Tools like web scrapers are becoming increasingly common for compliance functions, making it possible to navigate the wealth of data available on line and generate relevant insights. In Italy, for example, a regional environmental protection agency used an automated web scraper to identify thousands of businesses that had not applied for required licences that allow the agency to monitor pollution activities. The programme used public search engines like Google and Bing to identify businesses’ web pages (e.g. searching “car repair Trentino”), then compared business identification numbers from web pages with the list of licenced operators.

About one-third of OECD Members reported applying data-driven methods to monitor the impacts of laws and regulations. A similar proportion of Members reported applying data-driven methods to enforcement. Most Members who report applying data-driven methods note that they have adopted the practice relatively recently. Examples from various countries showcase how data analysis technologies can be employed to inform better monitoring and enforcement (Box 4.9).

There is significant room for greater adoption of data-driven tools and technology to enhance monitoring and enforcement, but governments must also exercise caution in doing so as the effectiveness of these tools depends on the quality of the data used. This is known as the “garbage in, garbage out” problem and poses a risk for data-driven regulatory delivery. The United States Securities Exchange Commission (SEC), responsible for protecting against market manipulation, uses machine learning to detect insider trading. However, two of the tools it uses for doing so are algorithms trained using data that were collected in connection with the SEC’s enforcement activities. These data, therefore, reflect the SEC’s judgements about the likelihood of market misconduct in each case. Consequently, “the types of misconduct and entities targeted [by the algorithm] will reflect the assumptions, heuristics, and biases of enforcement staff” (Allen, 2023[54]). This leaves the algorithm vulnerable to missing novel or more creative forms of insider trading. The SEC is working to implement a tool that would track all trading activity; training the algorithms using this significantly broader data set could improve the effectiveness of the algorithms (Allen, 2023[54]).

Importantly, digital tools can empower people to promote compliance and buy-in. During the COVID‑19 pandemic, for instance, smartphone applications were used to inform people about restrictions in place at a given time and area, and to support track-and-trace activities (UK Health Security Agency, 2022[55]). Some countries have developed tools for users to report concerns and get direct support. In Lithuania, as part of a plan to tackle rising waste production and littering in the countryside, in 2023, the environmental protection regulator launched a web and mobile application called “I manage Lithuania” (Tvarkau Lietuva) that enables citizens to report illegal waste (Lithuanian Ministry of Environment, 2024[56]). The regulator can then follow up to communicate when the report is received and addressed and provide feedback on the actions taken. This application was inspired by systems used in many cities to let people notify the municipal authorities of damaged public goods or areas in need of cleaning. The system offers several advantages: it is very simple to use and entirely transparent, as every message is public, showcasing the authority’s reactivity.

To proactively prevent non-compliance, other countries have developed web services to give users a better understanding of their obligations. As part of its food safety strategy, the Campania region of Italy has launched a self-assessment tool to reinforce business compliance, called GISA Self-assessment (Autovalutazione (Region of Campania, n.d.[57])). The web-based application allows companies to access and fill out the official and relevant inspection checklist autonomously. This enables them to identify the strengths and weaknesses of their facilities from the health authorities’ perspective. The tool is also available to “guest users”, enabling individuals to learn more about food and veterinary requirements before setting up a business, depending on its future characteristics. The result of the self-simulated inspection is expressed as a risk level. The purpose of GISA Autovalutazione is to educate; it is, therefore, not a mandatory self-monitoring or self-reporting tool. Advances in the use of predictive techniques may in the future allow the tool to also indicate the occurrence of potential risks to profiled companies by enhancing their historical data.

This dialogue not only supports compliance, it also promotes collaboration between people and regulators. Done properly, this can enhance the perceived legitimacy of rules and bolster public trust in government. For these tools to be adopted, the government must ensure a clean, simple and efficient implementation. Important success factors include:

• use of a non-proprietary platform

• guarantee security and privacy, often by allowing anonymous contributions

• automation when handling simpler cases, reserving human actions only for complex cases (Welby and Hui Yan Tan, 2022[58]).

By becoming “active users” of new technologies, governments can create better-informed rules and streamline regulatory delivery. Ultimately, this can lead to better compliance and better outcomes, from reduced regulatory burdens to enhanced knowledge and protections (OECD, 2020[45]). Box 4.10 provides some examples from economic regulators, who are taking advantage of digital technologies to be world-class regulators.

Rule-making is generally made under uncertainty (see Chapter 5). When a new digital product emerges, policymakers may not have the information they need to assess how the product will impact people, the market or other rules that are already in place. In these cases, regulatory experimentation6 gives policymakers a way to gather evidence to take well-informed decisions that support innovation without compromising protections and policy goals (OECD, 2021[11]). Policy experimentation can be furthered, enabling analysis of data that is in alignment with risk-, outcome- and performance-based design and delivery approaches to regulating digital innovations (OECD, 2018[25]; Attrey, Lesher and Lomax, 2020[60]).

Fundamentally, a regulatory experiment involves limited testing of a new regulatory approach to see how it works in practice, as opposed to implementing a rule based on guesswork or preventing innovation from entering the marketplace altogether. A regulatory sandbox more specifically is also an experiment, but normally characterised by some controlled departure from the existing regulatory framework within a defined space and time, and under the supervision of regulators; this departure can be, for example, a waiver from an existing rule, additional custom rules or a change in how rules are enforced within the sandbox. The lessons learnt can then inform how rules need to adapt and/or how an innovation needs to adapt to support economic growth without compromising health, safety or well-being. The types of experimentation are explored in further detail and with important nuance in OECD (2024[14]), and specifically in the case of AI in OECD (2023[61]).

There is a growing awareness among OECD Members about the potential and value of regulatory experimentation. Several jurisdictions are using regulatory sandboxes of varying scales to test new technologies (Attrey, Lesher and Lomax, 2020[60]). Box 4.11 discusses examples of regulatory sandboxes that have yielded tangible outcomes, with a focus on helping to safely bring new technologies to market.

However, there are limitations and costs associated with experimentation, sometimes requiring significant investments in time and resources. A successful experiment often requires a sufficient level of co‑ordination between sectoral regulators, as they deal with many cross-sectoral emerging and new technologies. In Korea, for example, the Regulatory Sandbox programme involves multiple ministries in the process while in Germany, the overarching Regulatory Sandbox Strategy is designed to cut across sectors and ministries (Attrey, Lesher and Lomax, 2020[60]). Looking from the participants’ side, the businesses involved in an experiment may have an advantage in the marketplace from getting a temporary derogation from the existing rules or gaining an enhanced understanding of how to navigate existing and new rules informed by the experiment. To some extent, these risks may be partially mitigated through careful planning and strategies for how data will be collected, analysed and actioned (OECD, 2024[14]).

Critically, to maximise the benefits of an experiment, policymakers must close the feedback loop by translating the evidence gathered into policy impact. This requires correctly interpreting the collected data, then having mechanisms to incorporate this evidence into the decision-making process to shape future rules as appropriate (OECD, 2024[14]). Estonia’s framework for public sector experimentation, for example, mentions potentially embedding testing as a tool for developing impact assessments. As more and more policymakers start using experimentation as a tool, sharing lessons learnt and best practices – particularly for integrating learnings from experiments into the rule-making process – can help promote the effectiveness of experimentation as a regulatory tool.

Innovation and its impacts cross both sectors and borders. Accordingly, recent initiatives have also explored the development of multi-jurisdiction regulatory experiments or sandboxes. These initiatives bring together policymakers and innovators from different areas – both subject and geographic – to test how regulation can help innovators to scale and operate safely across multiple jurisdictions. Outcomes could also include helping jurisdictions align their rules and address loopholes that could be exploited by firms.

In 2022, for example, the Global Financial Innovation Network published a report setting out learnings from its first live cross-border tests within their global sandbox. These tests provided practical insights of how innovative financial products and services operate in multiple markets. The sandbox also fostered collaboration between innovators and regulators across these markets, which continued beyond the tests (Global Financial Innovation Network, 2022[66]). The next section further discusses the importance of international co-operation in responding to innovation.

Regulatory institutions are often ill-equipped to address the regulatory challenges created by digital technologies or to implement the reforms required to set governing institutions up for success. As digital technologies increasingly blur traditional sectoral and jurisdictional lines, they expose the limitations of existing regulatory resources, skills and practices. It is critical that not only regulatory regimes, but the institutions that support them, be sufficiently supported to effectively address the multifaceted impacts of digital technologies – ensuring they are governed in a way that promotes innovation, protects public interests and upholds legal standards across diverse regulatory landscapes. Investing in regulatory institutions’ co-operation and capacity creates a more unified, cohesive, responsive regulatory environment.

New forms of digital technologies cut across and transform traditional sectors and markets, requiring strong co-ordination and concerted effort across government. In what is known as “technology convergence”, advances in one area can have an impact on and applications in another area. Therefore, new products and services like IoT devices, augmented reality applications and AI can be subject to policies and guidance from a myriad of different regulatory bodies. As a result, innovators may struggle to navigate the system and make sense of different or even conflicting advice and guidance. In other cases, digital innovations could “fall through the cracks” where institutional responsibility is unclear, creating a lack of effective oversight over new digital technologies.

In response and to better address the cross-cutting nature of digital innovation, OECD Members are taking steps to foster effective co-ordination across the administration. Approximately 40% of OECD Members declared that their ministries and regulatory agencies co-ordinate to identify and address the above-mentioned issues where different bodies share responsibility in an area of innovation. This could, for instance, entail mechanisms to provide joined-up advice to innovators based on an agreed-upon and consistent policy position. As shown in Figure 4.2, this co-ordination often involves joined-up regulation to enhance regulatory coherence, including across national and subnational levels of government.

Co-ordination mechanisms are an increasingly important tool to make joined-up approaches work in practice. Governments are already moving towards several different models that can be deployed. Formal co-ordination mechanisms include, but are not limited to, Australia’s Digital Platform Regulators Forum, Canada’s Digital Regulators Forum, Ireland’s Digital Regulators Group and the United Kingdom’s Digital Regulation Co-operation Forum. Importantly, such domestic co-ordination efforts can act as a foundation for broader international regulatory co-operation, helping to build a shared understanding and create common regulatory practices that can be extended across borders. Such activities have spawned the International Network for Digital Regulation Cooperation, where members from these networks come together to help build international relationships, gather insights from other jurisdictions and enable co-operation, furthering the transition from domestic collaboration to international regulatory frameworks.

Countries can also use less formal knowledge hubs, such as Israel’s Knowledge Hub on AI that serves as a repository for guidance and information for all government entities to access. Ad hoc approaches also exist, such as regulators collaborating on common studies or striking taskforces of experts from various departments. For example, France’s Centre of Expertise for Digital Platform Regulation is an interdepartmental office to understand how online platforms work and set or adapt regulation. The domestic experiences and insights gained from such mechanisms provide valuable contributions to the international regulatory landscape that others can learn from. Additional examples can be found in the recent OECD policy papers “Shaping a rights-oriented digital transformation” (OECD, 2024[67]) and “The intersection between competition and data privacy” (OECD, 2024[68]).

However, co-ordination may not solve regulatory gaps in which no institution has a mandate. In these instances, governments may choose to establish new regulatory institutions or allocate more power to existing ones. Such decisions need to balance the autonomy of newly created bodies with the necessity of co-operation within the existing institutional framework. For instance, while Spain created a separate agency for supervising AI, France and the Netherlands opted to establish supervision units within their data protection authorities (OECD, 2024[67]).

Box 4.12 presents several additional examples of how ministries, regulators and other stakeholders can collaborate to take an agile approach in response to cross-cutting innovation, including digital technologies.

To manage the largely global impacts of digital technologies, policymakers need to look beyond their borders to avoid fragmentation and loopholes (OECD, 2021[74]). Where reasonable and relevant, international co-operation should seek to align regulatory approaches across jurisdictions (OECD, 2021[11]). Coherence across jurisdictions can make it easier for positive digital technologies to scale internationally and, therefore, help improve economic outcomes. More importantly, consistency helps to implement and enforce rules in an interconnected world. Rules to facilitate information sharing across borders, for instance, can prevent digital banks from exploiting siloed information to help individuals evade sanctions across jurisdictions (Europol, 2023[75]).

Collaboration among international experts and regulatory practitioners is essential for developing a common evidence base of relevant approaches and best practices. The OECD Recommendation of the Council on International Regulatory Cooperation (OECD, 2022[76]) underscores the importance of such collaboration, advocating for enhanced co-ordination and co-operation among countries to address shared regulatory challenges. In all cases, as policymakers around the world face common challenges associated with new technologies, they need to learn from each other’s successes and failures (Box 4.13).

Institutions are the enabling entities through which regulatory policy draws its legitimacy. Governments need to invest in building strong institutional capacity to effectively manage digital technologies. Nonetheless, international conversations with regulatory agencies highlight concerns about institutions’ preparedness to deliver their important future roles in supervising and enforcing digital regulation. OECD Members are continuing to address challenges by focusing on their institutional frameworks, resourcing, skills and expertise.

To build capacity within government to effectively regulate in the digital age, governments will need to adapt their institutional settings and working methods. For example, across EU member states, the implementation of three major regulations – the Digital Services Act, Digital Markets Act and AI Act – have added new mandates, functions and powers to regulate in the digital sphere that must be implemented at the country level, in co-ordination with the European Commission. While these regulations seek to empower regulators to address many of the challenges noted above, building governments’ capacity enables governments to wield this power effectively.

National digital strategies are a foundational pillar in establishing governments’ capacity to regulate in the digital age. Countries are choosing different bodies to implement these strategies, balancing notions such as legitimacy, political power or the possibility for co-ordination. Austria, for instance, has allocated strategic responsibility for developing and co-ordinating a national digital strategy to a ministry dedicated to digital affairs. Meanwhile, other countries have allocated responsibility above ministerial level, for example to the Chancellery, the Prime Minister’s Office or the Presidency, such as Australia and Colombia (Gierten and Lesher, 2022[82]).

Second, regulators’ mandates, powers and legal systems may require reform in order to align with new regulatory structures and evolving sector needs reforms (OECD, 2020[45]). This includes internal structures, outdated administrational processes, resourcing, and skills and change management strategies for organisational culture to adapt to and support these new responsibilities. In Canada, the Annual Regulatory Modernization Bill prioritises addressing legal barriers to digitalising regulatory systems, such as requirements in law that enforce paper-based applications or reporting. In Germany, model language is being developed for experimentation clause provisions authorising regulatory experimentation under new and existing laws. These provisions have been implemented in the areas of autonomous driving, passenger transport, drones and digital identity.

Third, central oversight, co-ordination and advice can overcome silos and fragmentation while offering a way to pool resources, including staff. For instance, approximately half of OECD Members reported having a dedicated body dealing with innovation-friendly regulation, which includes issuing guidance and helping policymakers across government consider the impacts of regulation on innovation, including digital technologies. Research on regulatory approaches to AI note a similar trend, with all regulations in the sample including oversight mechanisms that foster co-ordination and guidance.

Without adequate resources, institutions designing policies and regulation and overseeing digital technologies will have little power to shape a positive digital landscape. Incorporating the processes and tools highlighted earlier in this chapter requires additional institutional capacity to adapt traditional systems. However, regulators who are already stretched thin in their day-to-day duties may not be able to spare the time and resources needed to try something new. In acknowledgement of this, some OECD Members provide incentives, including financial support to encourage the adoption of innovative approaches to regulatory policy and governance.

In these cases, incentives and institutional support can help convey a clear signal about the importance of embedding agile regulatory approaches into the governance agenda. This signal needs to reach across the administration, including independent regulators, and different levels of government. Box 4.14 presents selected examples of mechanisms governments have established to provide resourcing support for more agile regulation.

The novelty of innovation and agile regulation can create the need for new technical and skills. For instance, policy teams in government departments or regulators may require in-house expertise on how to design and implement a regulatory experiment for new digital technologies. However, a survey of 57 regulators on staffing and funding arrangements highlights that more than half have difficulties hiring well-qualified staff, especially in digital domains (OECD, 2022[86]). A lack of understanding of the available tools, risks and best practices can be a barrier to agile regulation.

It is, therefore, important that policymakers leading the way on agile regulation document and share their knowledge to build practical skills. The European Commission has a comprehensive Better Regulation Toolbox, which includes practical discussion of tools, including and beyond experimentation, to leverage the potential of innovation and reduce potential negative impacts (European Commission, 2017[87]).

Similarly, regulators may lack staff with in-depth technical skills, such as data scientists, to regulate complex technologies or make the best use of them to regulate more efficiently. Compared to digital firms, which can attract and pay the best and the brightest, regulators often have less competitive salaries and, in some cases, may lack the necessary funds. In some cases, regulatory agencies are collaborating to recruit relevant experts, for example by hiring them into a shared pool from which they can be surged into different agencies to help manage costs and provide an appealing workplace environment.

To overcome the gap in technical expertise for designing and administering data-driven regulatory methods, specialised centres of expertise can offer a solution. In 2020, Spain created the Data Office, whose competencies notably include:

• operating a Centre for Advanced Analysis of Data that will define the methodologies and best practices for decision making tools based on public sector data

• designing strategies for data management, and sharing among enterprises, citizens and public administrations

• defining public governance policies and standards for data management

• creating tools for knowledge transfer in the public administration.

  The Data Office’s mission is to boost the management, sharing and use of data throughout the different productive sectors of the Spanish economy and society.

By investing in and providing technical guidance, tools and training, government agencies can be better equipped with the knowledge and resources necessary to effectively manage and regulate emerging digital technologies. By centralising expertise, these centres can enhance the capacity of multiple departments and streamline technical regulatory processes.

[54] Allen, H. (2023), “Regulatory innovation and permission to fail: The case of Suptech”, New York University Journal of Law & Business, Vol. 19/2, pp. 237-308, https://digitalcommons.wcl.american.edu/facsch_lawrev/2216.

[46] Amaral, M. and G. Hernández (2020), Survey on Experiences with Regulatory Impact Assessments Related to Emerging Technologies: Overview of Results, OECD, Paris.

[60] Attrey, A., M. Lesher and C. Lomax (2020), “The role of sandboxes in promoting flexibility and innovation in the digital age”, OECD Going Digital Toolkit Notes, No. 2, OECD Publishing, Paris, https://doi.org/10.1787/cdf5ed45-en.

[52] Canadian Food Inspection Agency (CFIA) (2024), Risk Assessment models, https://inspection.canada.ca/en/about-cfia/cfia-2025/era-models.

[63] Challenge Works (2021), Regulators’ Experimentation Toolbox: Is a regulatory sandbox right for you?, https://issuu.com/nestachallenges/docs/cri_regulators_experimentation_toolkit_v8/s/13222074.

[7] Chen, Z. (2023), “Ethics and discrimination in artificial intelligence-enabled recruitment practices”, Humanities and Social Sciences Communications, Vol. 10/1, https://doi.org/10.1057/s41599-023-02079-x.

[35] Council of Europe (2024), Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law, Council of Europe, https://rm.coe.int/1680afae3c (accessed on 6 September 2024).

[69] Danish Business Authority (n.d.), About the DBA, https://danishbusinessauthority.dk/about-the-danish-business-authority.

[40] Département fédéral de l’économie, de la formation et de la recherche de la Suisse (2024), Manuel sur l’analyse d’impact de la réglementation (AIR), https://www.seco.admin.ch/dam/seco/fr/dokumente/Wirtschaft/Wirtschaftspolitik/Regulierung/handbuch_rfa.pdf.download.pdf/handbuch_rfa.pdf.

[51] Department of Transport of Estonia (2021), ITS development plan 2021-2025, https://www.transpordiamet.ee/its-arengukava-2021-2025.

[64] Droniq (2023), U-Space Realla-bor Hamburg, https://droniq.de/projekte/u-space-reallabor-hamburg/.

[32] European Commission (2024), AI Pact, https://digital-strategy.ec.europa.eu/en/policies/ai-pact (accessed on 6 September 2024).

[78] European Commission (2022), “Launch event for the Spanish Regulatory Sandbox on Artificial Intelligence”, https://digital-strategy.ec.europa.eu/en/events/launch-event-spanish-regulatory-sandbox-artificial-intelligence.

[87] European Commission (2017), Better Regulation “Toolbox”, European Commission.

[75] Europol (2023), European Financial and Economic Crime Threat Assessment 2023 – The Other Side of the Coin: An Analysis of Financial and Economic Crime, Publications Office of the European Union, Luxembourg.

[2] Evans, J. (20 January 2021), “A brave new world: Agile regulation to unleash innovation”, UK Civil Service Blog, https://civilservice.blog.gov.uk/2021/01/20/a-brave-new-world-agile-regulation-to-unleash-innovation (accessed on 18 March 2024).

[39] Federal Ministry of the Interior and Community of Germany (2017), What is the Online Access Act?, https://www.digitale-verwaltung.de/Webs/DV/EN/ozg/ozg-node.html.

[85] German Federal Ministry for Economic Affairs and Climate Action (2023), “Regulatory sandboxes – Testing environments for innovation and regulation”, web page, https://www.bmwk.de/Redaktion/EN/Dossier/regulatory-sandboxes.html.

[82] Gierten, D. and M. Lesher (2022), “Assessing national digital strategies and their governance”, OECD Digital Economy Papers, No. 324, OECD Publishing, Paris, https://doi.org/10.1787/baffceca-en.

[66] Global Financial Innovation Network (2022), The Global Financial Innovation Network Cross-Border Testing Initiative: Cohort 1.0, Global Financial Innovation Network.

[62] Government of Canada (2024), Regulatory Sandbox, https://www.canada.ca/en/government/system/laws/developing-improving-federal-regulations/modernizing-regulations/regulatory-sandbox.html.

[43] Government of Canada (2023), “Targeted regulatory reviews”, web page, https://www.canada.ca/en/government/system/laws/developing-improving-federal-regulations/modernizing-regulations/targeted-regulatory-reviews.html.

[77] Government of Spain (2022), “AI regulatory and ethical framework”, https://espanadigital.gob.es/en/measure/ai-regulatory-and-ethical-framework.

[27] Guston, D. (2013), “Understanding ’anticipatory governance’”, Social Studies of Science, Vol. 44/2, https://doi.org/10.1177/0306312713508669.

[15] Hernández, G. and M. Amaral (2022), “Case studies on agile regulatory governance to harness innovation: Civilian drones and bio-solutions”, OECD Regulatory Policy Working Papers, No. 18, OECD Publishing, Paris, https://doi.org/10.1787/0fa5e0e6-en.

[44] HM Treasury (2023), “Pro-innovation Regulation of Technologies Review”, web page, https://www.gov.uk/government/collections/pro-innovation-regulation-of-technologies-review.

[56] Lithuanian Ministry of Environment (2024), Tvarkay Lietuva (“I manage Lithuania”), https://tvarkaulietuva.lt (accessed on  2024).

[5] Matasick, C., C. Alfonsi and A. Bellantoni (2020), “Governance responses to disinformation: How open government principles can inform policy options”, OECD Working Papers on Public Governance, No. 39, OECD Publishing, Paris, https://doi.org/10.1787/d6237c85-en.

[29] McKinsey & Company (2023), Quantum Technology Monitor, https://www.mckinsey.com/~/media/mckinsey/business%20functions/mckinsey%20digital/our%20insights/quantum%20technology%20sees%20record%20investments%20progress%20on%20talent%20gap/quantum-technology-monitor-april-2023.pdf.

[80] Merchant, A. (2023), “Three recommendations for more inclusive and equitable AI in the public sector”, Open Government Partnership, https://www.opengovpartnership.org/stories/three-recommendations-for-more-inclusive-and-equitable-ai-in-the-public-sector.

[42] Ministry for Regulation (2024), “Regulatory stewardship”, web page, https://www.regulation.govt.nz/our-work/regulatory-stewardship.

[38] Ministry of Economic Affairs and Employment of Finland (2022), Innovation Impact Assessment: Guide for Law Drafting, https://julkaisut.valtioneuvosto.fi/bitstream/handle/10024/164443/MEAE_guidelines_16_2022_Assessment_regulation_innovation_15112022.pdf?sequence=1&isAllowed=y.

[49] Ministry of Economy of Latvia (2017), Guidelines on the application of the “Consult first” principle in the work of state institutions, https://www.em.gov.lv/sites/em/files/2017-06-15_10_05_37_kv_vadlinijas_final_makets_14.06.20171_0.pdf.

[65] Ministry of Health of Singapore (2023), Licensing experimentation and adaptation programme - a MoH regulatory sandbox, https://www.moh.gov.sg/sitefinity/status?ReturnUrl=http%3a%2f%2fwww.moh.gov.sg%2fhome%2four-healthcare-system%2flicensing-experimentation-and-adaptation-programme-(leap)---a-moh-regulatory-sandbox.

[48] Murray–Darling Basin Authority (2023), “Water resource plans”, web page, https://www.mdba.gov.au/water-management/basin-plan/water-resource-plans.

[37] National Regulatory Control Council (2022), Reducing Bureaucracy in a Changing World: Annual Report 2022, National Regulatory Control Council, Berlin, https://www.normenkontrollrat.bund.de/Webs/NKR/SharedDocs/Downloads/EN/Annual-Reports/Annual-Report-2022.pdf?__blob=publicationFile&v=2.

[73] New Zealand Customs Service (2024), Joint Border Analytics, https://www.customs.govt.nz/about-us/your-information/customs-information/joint-border-analytics/.

[6] OECD (2024), Facts Not Fakes: Tackling Disinformation, Strengthening Information Integrity, OECD Publishing, Paris, https://doi.org/10.1787/d909ff7a-en.

[1] OECD (2024), “Framework for Anticipatory Governance of Emerging Technologies”, OECD Science, Technology and Industry Policy Papers, No. 165, OECD Publishing, Paris, https://doi.org/10.1787/0248ead5-en.

[34] OECD (2024), “OECD launches pilot to monitor application of G7 code of conduct on advanced AI development”, press release, https://www.oecd.org/en/about/news/press-releases/2024/07/oecd-launches-pilot-to-monitor-application-of-g7-code-of-conduct-on-advanced-ai-development.html (accessed on 6 September 2024).

[3] OECD (2024), OECD Survey on Drivers of Trust in Public Institutions – 2024 Results: Building Trust in a Complex Policy Environment, OECD Publishing, Paris, https://doi.org/10.1787/9a20554b-en.

[14] OECD (2024), “Regulatory experimentation: Moving ahead on the agile regulatory governance agenda”, OECD Public Governance Policy Papers, No. 47, OECD Publishing, Paris, https://doi.org/10.1787/f193910c-en.

[33] OECD (2024), Report on the Implementation of the OECD Recommendation on Artificial Intelligence, OECD, Paris, https://one.oecd.org/document/C/MIN(2024)17/en/pdf.

[67] OECD (2024), “Shaping a rights-oriented digital transformation”, OECD Digital Economy Papers, No. 368, OECD Publishing, Paris, https://doi.org/10.1787/86ee84e2-en.

[68] OECD (2024), “The intersection between competition and data privacy”, OECD Roundtables on Competition Policy Papers, No. 310, OECD Publishing, Paris, https://doi.org/10.1787/0dd065a3-en.

[17] OECD (2023), G7 Hiroshima Process on Generative Artificial Intelligence (AI): Towards a G7 Common Understanding on Generative AI, OECD Publishing, Paris, https://doi.org/10.1787/bf3c0c60-en.

[61] OECD (2023), “Regulatory sandboxes in artificial intelligence”, OECD Digital Economy Papers, No. 356, OECD Publishing, Paris, https://doi.org/10.1787/8f80a0e6-en.

[16] OECD (2023), “The state of implementation of the OECD AI Principles four years on”, OECD Artificial Intelligence Papers, No. 3, OECD Publishing, Paris, https://doi.org/10.1787/835641c9-en.

[22] OECD (2022), Anticipatory Innovation Governance Model in Finland: Towards a New Way of Governing, OECD Publishing, Paris, https://doi.org/10.1787/a20998d5-en.

[86] OECD (2022), Equipping Agile and Autonomous Regulators, The Governance of Regulators, OECD Publishing, Paris, https://doi.org/10.1787/7dcb34c8-en.

[76] OECD (2022), Recommendation of the Council on International Regulatory Co-operation to Tackle Global Challenges, OECD, Paris, https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0475#dates.

[23] OECD (2021), Futures of Public Administration: Scenarios for Talent Management in Slovenia, OECD, Paris, https://oecd-opsi.org/wp-content/uploads/2021/10/Slovenia_Talent_Management_Scenarios_Final.pdf.

[74] OECD (2021), International Regulatory Co-operation, OECD Best Practice Principles for Regulatory Policy, OECD Publishing, Paris, https://doi.org/10.1787/5b28b589-en.

[31] OECD (2021), Practical Guidance on Agile Regulatory Governance to Harness Innovation, OECD, Paris, https://legalinstruments.oecd.org/public/doc/669/9110a3d9-3bab-48ca-9f1f-4ab6f2201ad9.pdf.

[11] OECD (2021), Recommendation of the Council for Agile Regulatory Governance to Harness Innovation, OECD, Paris, https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0464.

[8] OECD (2021), Regulatory Policy Outlook 2021, OECD Publishing, Paris, https://doi.org/10.1787/38b0fdb1-en.

[24] OECD (2021), Towards a Strategic Foresight System in Ireland, OECD Policy Brief, OECD, Paris, https://oecd-opsi.org/publications/foresight-ireland.

[13] OECD (2020), “Going Digital integrated policy framework”, OECD Digital Economy Papers, No. 292, OECD Publishing, Paris, https://doi.org/10.1787/dc930adc-en.

[30] OECD (2020), Regulatory Policy in the Slovak Republic: Towards Future-Proof Regulation, OECD Reviews of Regulatory Reform, OECD Publishing, Paris, https://doi.org/10.1787/ce95a880-en.

[45] OECD (2020), Shaping the Future of Regulators: The Impact of Emerging Technologies on Economic Regulators, The Governance of Regulators, OECD Publishing, Paris, https://doi.org/10.1787/db481aa3-en.

[72] OECD (2020), “Sperimentazione Italia – Observatory of Public Sector Innovation”, web page, https://oecd-opsi.org/innovations/sperimentazione-italia.

[9] OECD (2019), Recommendation of the Council on Artificial Intelligence, OECD, Paris, https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0449.

[10] OECD (2019), Recommendation of the Council on Responsible Innovation in Neurotechnology, OECD, Paris, https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0457.

[25] OECD (2018), OECD Regulatory Policy Outlook 2018, OECD Publishing, Paris, https://doi.org/10.1787/9789264303072-en.

[26] OECD (2015), OECD Regulatory Policy Outlook 2015, OECD Publishing, Paris, https://doi.org/10.1787/9789264238770-en.

[18] OECD (n.d.), “AI risk & accountability”, web page, https://oecd.ai/en/site/risk-accountability/publications (accessed on 27 September 2024).

[12] OECD (n.d.), “Global Forum on Technology”, web page, https://www.oecd.org/en/networks/global-forum-on-technology.html (accessed on 13 August 2024).

[59] OECD (forthcoming), “Paper on AI in regulatory delivery”, OECD, Paris, forthcoming.

[20] OECD (Forthcoming), Regulatory governance approaches to AI policy.

[19] Plonk, A., K. Perset and S. Fialho Esposito (23 July 2024), “OECD pilot on monitoring the application of the G7 Code of Conduct for Organisations Developing Advanced AI Development”, AI Wonk Blog, https://oecd.ai/en/wonk/pilot-g7-monitoring (accessed on 274 September 2024).

[57] Region of Campania (n.d.), “GISA Autovalutazione”, web page, https://www.regione.campania.it/regione/it/la-tua-campania/servizi-af3a/gisa-autovalutazione (accessed on 26 June 2024).

[28] Regulatory Horizons Council (2024), Regulating Quantum Technology Applications, https://assets.publishing.service.gov.uk/media/65ddc83bcf7eb10015f57f9f/RHC_regulation_of_quantum_technology_applications.pdf.

[70] Regulatory Horizons Council (2023), “Regulatory Horizons Council”, https://www.gov.uk/government/groups/regulatory-horizons-council-rhc#reports.

[71] Riigikantselei (2022), Framework for Public Sector Experimentation, https://riigikantselei.ee/media/2007/download.

[4] Ryan-Mosley, T. (2022), “The world is moving closer to a new cold war fought with authoritarian tech”, MIT Technology Review, https://www.technologyreview.com/2022/09/22/1059823/cold-war-authoritarian-tech-china-iran-sco/ (accessed on 13 August 2024).

[81] Standards Australia (2022), Critical and Emerging Technologies in South-East Asia, https://www.standards.org.au/engagement-events/international/critical-emerging-technologies.

[50] Statistics Estonia (2020), Mobility analysis, https://www.stat.ee/en/avasta-statistikat/eksperimentaalstatistika/mobility-analysis.

[53] Swiss Financial Market Supervisory Authority (2021), Artificial intelligence in the Swiss financial market, https://www.finma.ch/en/documentation/dossier/dossier-fintech/kuenstliche-intelligenz-im-schweizer-finanzmarkt/.

[21] Tõnurist, P. and A. Hanson (2020), “Anticipatory innovation governance: Shaping the future through proactive policy making”, OECD Working Papers on Public Governance, No. 44, OECD Publishing, Paris, https://doi.org/10.1787/cce14d80-en.

[83] UK Government (2022), “Notice: Projects selected for the Regulators’ Pioneer Fund (2022)”, web page, https://www.gov.uk/government/publications/projects-selected-for-the-regulators-pioneer-fund/projects-selected-for-the-regulators-pioneer-fund-2022.

[55] UK Health Security Agency (2022), “NHS COVID-19 app”, https://www.gov.uk/government/collections/nhs-covid-19-app.

[79] UNESCO (2022), Recommendation on the Ethics of Artificial Intelligence, United Nations Educational, Scientific and Cultural Organization, Paris, https://unesdoc.unesco.org/ark:/48223/pf0000381137.

[36] United Nations (n.d.), Global Digital Compact, https://www.un.org/techenvoy/global-digital-compact (accessed on 6 September 2024).

[41] United States Office of Management and Budget (1998), CIRCULAR NO. A-119 Revised, https://www.whitehouse.gov/wp-content/uploads/2017/11/Circular-119-1.pdf.

[47] Walprecht, S. and C. Lewerenz (2024), “Facilitating regulatory impact assessments: The benefits of machine learning in legislation”, presentation, https://www.destatis.de/EN/About-Us/Events/Machine-Learning/Slides/s2_walprecht.pdf?__blob=publicationFile.

[58] Welby, B. and E. Hui Yan Tan (2022), “Designing and delivering public services in the digital age”, OECD Going Digital Toolkit Notes, No. 22, OECD Publishing, Paris, https://doi.org/10.1787/e056ef99-en.

[84] World Economic Forum (2020), Agile Regulation for the Fourth Industrial Revolution: A Toolkit for Regulators, World Economic Forum, Geneva, https://www.weforum.org/about/agile-regulation-for-the-fourth-industrial-revolution-a-toolkit-for-regulators.