Assessing the accuracy and completeness of taxpayer reported information is a core function of tax administrations. This chapter takes a closer look at tax administrations’ work in this area, including what they do to understand and manage compliance risk, and how they prevent non-compliance. Finally, it looks at tax administration performance in addressing non-compliance.
Tax Administration 2025
Comparative Information on OECD and other Advanced and Emerging Economies
Abstract
The audit, verification and investigation function assesses the accuracy and completeness of taxpayer reported information. This function employs on average close to 30% of tax administration staff with their core task being to assure that tax obligations have been met. While this often happens through audits, there is an increasing use of automated electronic checks, validations and matching of taxpayer information. The undertaking and visibility of these and other compliance actions is critical in supporting voluntary compliance, including through their impacts on perceptions of fairness in the tax system, as well as creating a ‘deterrent effect’. This chapter therefore looks at:
How tax administrations manage compliance risks, including the different approaches to prevent non-compliance; and
The delivery of compliance actions undertaken by tax administrations, looking at electronic compliance checks, audits as well as tax crime investigations.
The process of compliance risk management, as described in the 2004 OECD guidance note Compliance Risk Management: Managing and Improving Tax Compliance (OECD, 2004[1]), has remained largely unchanged over the years. Its key steps, as illustrated in Figure 6.1., still serve as a blueprint for managing compliance risks. Since then, several OECD reports explored aspects of this framework providing guidance and good practice examples, and the 2017 report The Changing Tax Compliance Environment and the Role of Audit (OECD, 2017[2]) looked at a range of incremental changes occurring across tax administrations which, taken together, were changing the nature of the tax compliance environment, allowing for more targeted and managed compliance.
Source: OECD (2004), Compliance Risk Management: Managing and Improving Tax Compliance, https://www.oecd.org/content/dam/oecd/en/topics/policy-issues/tax-administration/compliance-risk-management-managing-and-improving-tax-compliance.pdf (accessed 1 October 2025).
This section examines how tax administrations are organising their processes in this area by looking at tax administrations approaches towards understanding and managing compliance risks, and some of the steps taken by administrations as regards preventing non-compliance.
Understanding and managing tax compliance risks
Understanding tax compliance risks is crucial for tax administrations as it enables them to identify areas where non-compliance is most likely to occur, whether due to error, negligence or deliberate action. By analysing and managing these risks, administrations can allocate resources more effectively, tailor enforcement strategies, and design programmes to improve voluntary compliance. As noted in the 2024 edition of this series, around 85% of administrations report having a formal compliance risk management strategy, with almost all of those having in place dedicated approaches for identifying, assessing and prioritising key compliance risks. (OECD, 2024[3])
Increasing availability of data
The huge increase in the amount of data available to tax administrations for compliance purposes is helping them to better understand risk areas and promote tax compliance. As more and more data is stored electronically, and the transfer, storage and integration of data has become easier through the application of new techniques and processes, tax administrations are now frequently tapping into data sources such as:
Data from devices: Data can be collected from devices that register transactions such as online cash registers and trip computers for taxis and trucks, and also gate registrations from barriers and weigh bridges. As Table 6.1. illustrates, slightly more than half of the administrations receive data from electronic fiscal devices or cash registers and in two-thirds of those situations, data is transferred automatically. A small number of administrations also receive data from other devices such as taxi meters.
Data from banks, merchants or payment intermediaries and service providers: This allows direct verification of income or assets reported by the taxpayer. Some jurisdictions receive transaction details or transaction totals for taxpayers on a regular basis.
Data from suppliers: Collecting data from suppliers, either directly or through the taxpayer, allows a more complete picture to be drawn about the activities and income of the taxpayer. This is seen in the increasing use of electronic invoicing systems (see Table 6.1.) which, as noted in Chapter 4, allows some tax administrations to prefill tax returns.
Data from customers: This is easiest in cases where the number of customers is limited and known, but increasingly mechanisms to leverage customer data are being used, for example in the verification of cash receipts.
Unstructured data concerning the taxpayer: Electronic traces relevant to business activities and transactions can be found on the internet and in social media.
Data from other government agencies: Data held by other government agencies for example for licencing, regulatory or social security purposes can be relevant in verifying tax returns or in risk assessments.
Data from international partners: International exchanges of data from the International Standards for Automatic Exchange of Information in Tax Matters (OECD, 2023[4]) and Country-by-Country Reporting (OECD, 2015[5]) is massively increasing the quantity of data available on international activity and providing useful information for audit and case selection processes and in some cases for prefilling of tax returns.
Percentage of administrations
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Tables A.105 Compliance approaches: Electronic invoicing, and A.106 Compliance approaches: Devices that register transactions, http://isoradata.org (accessed on 1 October 2025).
Mexico – Digitalisation of foreign trade procedures
One of the guiding principles of the Tax Administration Service’s (Servicio de Administración Tributaria, SAT) audit plan is the digitalisation of foreign trade notices, aiming to simplify processes and make it easier for taxpayers to submit foreign trade notices through the use of various technological tools.
This digital transition has allowed SAT to facilitate and simplify procedures. It saves time for taxpayers as they no longer have to travel to the SAT offices to submit notices in person and guarantees compliance with tax and customs obligations as submitting digitally ensures they are in the correct format. It also means that the SAT has better control of the information collected from the notices as they are stored digitally, which enables ease of access when doing analysis for audits. Finally, it benefits the environment by reducing the use of paper.
United Kingdom – Electronic trade documents
HMRC have made significant progress in digitalising trade processes, which is helpful due to a significant proportion of the data in electronic trade documents being required for tax processes. This also makes trade easier for UK businesses, especially small and medium enterprises, which in turn increases growth.
This therefore presents an opportunity for HMRC to access and use this data quicker to satisfy its requirements. This will not only improve HMRC’s risk and compliance processes, but also reduce the administrative burden on businesses. For example, a commercial invoice in the form of an Electronic Transferable Record (ETR) would assure VAT reporting. This offers a potential approach for electronic-invoicing (e-invoicing). Innovation could demonstrate the feasibility of treating an e-invoice in the same manner as an ETR and therefore, the principles of a ‘reliable system’ would apply with potential applications in VAT reporting.
In support of this, HMRC is identifying current difficulties for both internal and external users, to assess the value of electronic trade documents in relation to improving the design and administration of both tax and trade processes, for businesses and the government. To that effect, user research is underway for both external businesses as well as tax administration officials.
Sources: Mexico (2025) and the United Kingdom (2025).
With increasing amounts of data being handled by tax administrations, the implementation of mechanisms to protect and manage data is now commonplace, and a critical function. These mechanisms support wider data governance processes, and in turn help maintain taxpayer trust in the system as well as meet legal obligations (see Table 6.2.). Box 6.2. highlights a few examples in this area. Moreover, as data systems become more connected, the importance of cyber security is growing. The 2023 edition of this publication contained a few examples from tax administrations in this space – see Tax Administration 2023, Box 10.5. (OECD, 2023[6]).
Percentage of administrations that have the respective process in place
Source: OECD (2025), Tax Administration Digitalisation and Digital Transformation Initiatives, Table 4.1., https://doi.org/10.1787/c076d776-en.
Australia – Whole-of-government Data Ethics Framework
The whole-of-government Data Ethics Framework provides guidance for the Australian Public Service to safely and appropriately expand their use of data and analytics in new and innovative ways, while building the Australian community’s trust that their data is being used in ethical ways. The Framework identifies next steps for extending beyond current data uses as new technologies emerge and are adopted, which is relevant to the Australian Taxation Office (ATO) as it continues its data and digital transformation journey.
The Framework includes:
Data ethics principles;
Guidance on applying the principles to identify risks;
Roles and responsibilities;
Use cases to demonstrate where the Framework can be applied.
For further information on the Framework see here: https://www.finance.gov.au/government/public-data/public-data-policy/data-ethics-framework (accessed on 1 October 2025).
Norway – The data-driven future
The NTA is a data-driven organisation, and over the past years has boosted its effectiveness through data-driven taxpayer management initiatives. However, so far this has been on a relatively small scale. Therefore, to further improve its efficiency, the NTA is seeking to scale up its data-driven solutions.
To achieve this, a cloud-based data and analytics platform has been established, featuring capabilities for analytics, data science/machine learning, as well as dashboards and reporting. It is currently being utilised for a wide range of analytics projects and reporting use cases, from cryptocurrency to data quality and fraud detection. The NTA is also using the platform for the development and training of machine learning models and exploring generative Artificial Intelligence-based solutions with promising results, particularly in customer support and the improved handling of unstructured documents.
Effective data governance measures are key to scaling up data-driven solutions. These include establishing new roles, security mechanisms, and legal processes, as well as ensuring the data quality. The NTA makes analytical data available through data products, which are easily accessible datasets with a range of data, from taxpayer data to more administrative data. Nearly 200 data products have been produced so far, with many more to come. Recognising the scale of the data-driven change, an organisation-wide data literacy initiative is also underway to bring people in the organisation along with the changes and increase understanding.
Spain – Methodology for the development of artificial intelligence projects
AEAT has established a specific methodology to support the development of its AI projects, with the aim of improving the quality of the projects and enhancing their governance. The methodology has been specifically tailored to AI projects, addressing each phase of their lifecycle.
The methodology differs from a software development-oriented approach in that it does not focus on the technical aspects of code development, but rather on the various specific aspects that must be analysed in any AI project - such as data sources, biases and periodic retraining.
The methodology defines a series of stages and tasks, and specifies the roles and stakeholders in each of them. Therefore, most departments of the AEAT, including IT, business, security, and legal, can contribute their knowledge and perspectives to the AI project. All relevant issues related to the projects are analysed, including data quality, the existence of potential biases, transparency, and integration with corporate applications.
If the project fails to gain approval from all parties due to non-compliance with the ethical, legal, or quality standards established by the AEAT, it cannot proceed. Applying the methodology to a specific project ensures its quality and provides documentation that guarantees its traceability. The methodology will undergo periodic reviews, to adapt to new technological and legal situations that arise.
Sources: Australia (2025), Norway (2025) and Spain (2025).
Data analytics
Over recent years, the application of advanced analytics to risk management and risk targeting is becoming increasingly common:
Table 6.3. shows 87% of tax administrations reporting using big data in their work, and most of those that use big data are using it to improve their compliance work. Big data is also used for forecasting and a variety of other purposes as can be seen in Table 6.3. and the example in Box 6.3.
Of the 58 tax administrations covered by this report, nearly all report using data science / analytical tools with the remaining administrations in the process of preparing the use of such tools going forward (see Table 6.4.).
Similarly, the use of artificial intelligence (AI), including machine learning, for risk assessments and detecting fraud is already undertaken by the majority of administrations (see Table 6.3.).
Percentage of administrations
Source: OECD (2025), Tax Administration Digitalisation and Digital Transformation Initiatives, Table 4.11. and Table 5.6., https://doi.org/10.1787/c076d776-en.
Australia – Central Dashboard
The ATO has introduced Centrl, a dashboard that brings together data from multiple systems into one place to help staff working with public, private and multinational enterprises to manage their allocated cases. Developed by a cross-functional team from business and data areas, Centrl connects to the ATO’s case management, tax filing data store, and profiling systems.
ATO officials can use Centrl for various functions, including to:
Get a list of their cases at the taxpayer group level with key information displayed upfront;
Get alerts when entities in a taxpayer group have taken specific actions (for example, filing an income tax return) or been matched in new data (such as property transactions);
Post notes to other case participants, regardless of team or area, to share information and set due date reminders to assist in keeping the case on-track;
Access a news feed to stay informed about system updates, changes to procedures or certain risks to look out for in cases;
Navigate directly (via an application panel) to frequently used systems that will open with pre-loaded data for the selected taxpayer group or entity;
Search for a taxpayer by their name, Australian business number or tax file number;
Navigate directly (via footer links) to relevant corporate applications, tools and external resources.
Centrl saves staff from manually monitoring multiple systems for events relevant to their cases, providing an efficient way to collaborate with other staff involved in the engagement and receive targeted news related to their work. This gives them more time to spend on the interpretation and application of tax law.
China (People’s Republic of) – Applying tax data to serve national governance
The STA is fully leveraging tax-related data to deepen its tax and economic analysis which it can share with the wider government. This effort aims to provide a high-level service platform for the modernisation of the tax system and ability to effectively govern.
At the macro level, the STA has utilised data from electronic invoice accounting to make up a model which reflects macroeconomic trends such as time, tax category, industry, and region. This provides a solid foundation for the central government to formulate macroeconomic policies.
At the middle level, a tax comparative analysis system has been established. It can automatically conduct comparative analysis between different regions, cities and provinces, both domestically and internationally. This helps identify regions and groups with significant tax contributions and assists local government in understanding the areas’ economic development.
At the micro level, real-time aggregation of enterprise “one-household” data allows for a comprehensive view of business operations, tax payments, invoices and risks, as well as corporate financial statements. This provides valuable insights into the underlying patterns, trends and risks of economic activities.
In recent years, the STA has produced over ten thousand analytical reports annually, supporting national macroeconomic decision-making, social operation analysis, and assessments of business conditions.
Sources: Australia (2025) and China (People’s Republic of) (2025).
Percentage of administrations that use this technology
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Tables A.108 Innovative technologies: Implementation and usage - Blockchain, artificial intelligence, and cloud computing, and A.109 Innovative technologies: Implementation and usage - Data science, robotic process automation, and APIs, http://isoradata.org (accessed on 1 October 2025).
This increasingly sophisticated use of analytics on expanded data sets is leading to a sharpening of risk management and the development of a range of intervention actions, including through automated processes. Box 6.4. contains a variety of examples on the use of advanced analytics techniques by tax administrations.
Additionally, the OECD report Advanced Analytics for Tax Administration: Putting data to work (OECD, 2016[7]) provides practical guidance on how tax administrations can use analytics to support compliance and service delivery.
With the use of analytics becoming a common and integrated part of tax administrations across the world, in developed and developing countries alike, the OECD’s Forum on Tax Administration (FTA) developed the Analytics Maturity Model (OECD, 2022[8]) to help tax administrations self-assess their current level of maturity in their analytics usage and capability. This provides insight into their current status by identifying areas of weaknesses as well as strengths.
Australia – Unstructured data analytics
The ATO is developing a selection of tools to support faster, more reliable use and analysis of the large amounts of data it collects in unstructured formats, including AI models, robotic process automation and search functionality.
An example is a new tool used in the Justified Trust programme, which is used to assure the ATO’s top 1 000 public and multinational businesses that have substantial economic activity related to Australia. Following the completion of an assurance review, a comprehensive report is prepared outlining ATO’s findings and recommendations for each area that is assured. The unstructured nature of the information previously limited the ATO’s ability to develop intelligence across the hundreds of reports generated by the programme. However, the new analytics capability extracts information from the reports, using AI to identify information of interest and index the text. A user interface is available through which the information across the set of reports can be filtered and searched.
This enables case teams to better identify taxpayers with similar issues and more efficiently undertake future assurance reviews, as well as ensuring consistency across the population. This has improved the ATO’s reporting abilities, due to the ability to provide greater insights and granularity at the same time. This in turn enables the ATO to provide clearer guidance to the population through its annual Findings Report. Risk and intelligence functions are also better able to identify trends and emerging areas of concern, due to the increased visibility they have of the assurance outcomes.
Australia – Advanced Analytics Platform Cloud
The ATO has introduced the Advanced Analytics Platform Cloud (AAP Cloud), which provides a secure environment for data scientists and engineers to develop machine learning models. The AAP Cloud infrastructure is hosted in a private cloud and provides secure access to data in ATO systems.
Open-source services have been used, which deliver cost savings and have the flexibility to adapt when new technologies emerge, or requirements change. The use of cloud computing also provides the ability to dynamically scale operations to access additional computing power, which is an advantage over other options such as on-premises infrastructure.
The machine learning models operating on the AAP Cloud are being used to identify potential non-compliance, and the ability to develop and deploy artificial intelligence at scale has helped support the treatment of risks, particularly fraud, where risk behaviours have rapidly changed. The models have also supported tailored engagement with taxpayers and their representatives in real-time, helping the ATO in providing a better service to taxpayers.
China (People’s Republic of) – Integrating geographic information technology with tax management
In 2024, the STA launched the Taxation-Geographic Information System (T-GIS), a system that integrates geographic information technology with tax management. This system enhances the tax filing experience for taxpayers and facilitates the efficiency of tax risk management and decision-making for tax officials.
First, T-GIS transforms tax filing through visualisation. It has established a database with 38.5 million locations and numerous taxpayer addresses, which are integrated into the online E-Tax Platform. This enables taxpayers to have information automatically displayed on their real estate assets, the applicable tax rates, and even to measure land areas. As a result, the efficiency and accuracy of tax filing are significantly improved.
Second, by geocoding taxpayers' registered addresses and mapping housing and land tax sources, T-GIS allows tax officials to identify compliance risks remotely using AI and modelling comparison techniques. AI can automatically identify taxable lands in different forms and achieves an intelligent monitoring of tax sources from remote images, while a modelling comparison superimposes and compares various tax-related information on the map to detect potential tax evasion and avoidance risks.
Third, T-GIS equips tax officials with multi-dimensional analysis tools that cover nine tax categories, looking at a macro to micro perspective. Tax officials can carry out macro research on tax revenue, tax sources and economic analysis of specific taxes, as well as micro-level studies using satellite images and detailed tax source locations.
Italy – Innovations to counter Value Added Tax fraud
The IRA has taken innovative steps to fight against VAT fraud.
Since 2024, the IRA has been working on a project which aims to map ‘high intensity VAT evasion’, through advanced analytical techniques. The model is used to group taxpayers into clusters corresponding to different levels of risk. This then allows the IRA to predict potential VAT evasion by high-risk persons and prevent revenue loss.
A new tool has also been developed – TaxnetVA – that integrates varied data sources to enable a visualisation of the different relationships between economic entities. This makes it easier to identify any suspicious chains, which characterise intra-community VAT fraud. This information is presented in easily interpretable graphs. Economic entities are displayed as points (called ‘nodes’), and the relationships between them drawn as ‘edges’. Each node has a list of properties, such as business activity, turnover, legal form etc. Similarly, information is also given for each edge, such as transaction and VAT amount, shareholder percentage, and corporate role. The nodes and edges are characterised with different colours and sizes depending on their specific characteristics (such as active verses ceased activity or partnership verses company), to give a clear visual of how risky the enterprise is.
Japan – Using artificial intelligence and data analytics to collect tax
The NTA is working to refine and improve tax collection by using AI to analyse a wide range of data in determining how best to respond to non-compliant taxpayers based on their circumstances.
Specifically, through business analytics tools and programming languages, the NTA has built a model which can predict the most effective method of contacting each non-compliant taxpayer. This includes making phone calls, in-person visits and sending letters based on various personal information (including previous contact records, tax-return data and the taxpayer’s business type), so that taxes can be collected efficiently.
The NTA has also constructed a response prediction model that analyses information on non-compliant taxpayers, including their past contact records, which predicts the days of the week and time of day when they are more likely to respond to phone calls. Consequently, the response rate increased by 8.6 percentage points compared with non-use of AI, during the period of July 2023 to June 2024.
Thailand - Enhanced income and invoice classification
Thailand's tax administration has integrated NLP to innovate and streamline the classification of taxable income and electronic-invoicing (e-invoicing) data.
Previously, manual processing of withholding tax forms and invoices was time-intensive and prone to errors due to the varied and detailed descriptions provided by taxpayers. By implementing NLP, the accuracy of income classification in withholding tax data has improved by 60%, enabling better risk assessment. This system supports the detection of discrepancies, such as inconsistencies between reported income in withholding taxes versus VAT declarations; and assists in compliance analysis by identifying taxpayers who haven't filed annual returns, despite having tax withheld.
Moreover, NLP plays a crucial role in efforts to expand the tax base through the categorisation of e-invoicing data. By efficiently sorting invoices into personal and business-related expenditures, NLP simplifies the process of identifying businesses that are not registered in the tax system and ensuring they register. The results of this categorisation guide follow-up actions, such as notifying potential non-compliant entities. This integration enhances compliance and equips tax officers with a reliable tool for conducting thorough and efficient tax audits.
The methodology is based on modelling each phase of the process through a suitable probability distribution. These different models are then combined to predict a score expressing the overall profitability of the investigation of each selected taxpayer.
Sources: Australia (2025), China (People’s Republic of) (2025), Italy (2025), Japan (2025) and Thailand (2025).
Taxpayer programmes
Another approach for targeted risk management is the creation of units looking into the tax affairs of specific taxpayer segments. Two specific areas where tax administrations have found it advantageous to manage specific groups of taxpayers on a segmented basis are large business taxpayers, and high net wealth individuals (HNWIs). The rationale for focusing administration resources on managing these groups revolves around the:
Significance of tax compliance risks: due to the nature and type of transactions, offshore activities, opportunity and strategies to minimise tax liabilities; and in the case of large business, the differences between financial accounting profits and the profits computed for tax purposes.
Complexity of business and tax dealings: particularly the breadth of their business interests and in the case of HNWI, the mix of private and tax affairs.
Integrity of the tax system: the importance of being able to assure stakeholders about the work undertaken with these high-profile groups of taxpayers.
The SII has implemented a new risk assessment framework, introducing risk zone indicators to monitor multinational business groups, high-net-worth individuals, and strategic industries. This system aims to enhance transparency and compliance by evaluating the tax behaviour of taxpayers engaged in international operations or classified as high net worth individuals.
The framework establishes three risk levels—low, medium, and high—with specific metrics tailored to each category:
Mining business groups: The risk classification is based on four key metrics: effective business tax rate; overall effective tax rate; total tax contribution; and three-year tax burden. These metrics are evaluated using sector-specific interquartile ranges.
High-net-worth individuals: They undergo a two-phase assessment process. Initially, they are categorised based on total wealth. Subsequently, an effective tax rate is calculated considering all income sources, regardless of taxability, to determine their risk classification.
Industrial sectors involved in transfer pricing operations: Specifically motor vehicles, machinery, healthcare, and electronics—risk zones are determined by analysing the relationship between a company's functional profile, asset ownership, and risk assumption in distribution activities.
This comprehensive framework strengthens SII's ability to identify and manage tax compliance risks while providing taxpayers with clearer guidance on their risk status. By implementing these targeted assessment criteria, SII aims to promote more effective tax oversight and encourage greater compliance among international taxpayers.
Source: Chile (2025).
Additionally, in the case of large taxpayers, while being a small number of taxpayers, they are typically responsible for a disproportionate share of tax revenue collected. Even though large taxpayer offices/ programmes manage only 2.7% of corporate taxpayers, on average they account for 44% of all net revenue collected, including withholding payments on behalf of employees (Table 6.5.). Looking at the individual country-level, the data indicates that for most jurisdictions between 30% and 60% of their total net revenue was received from taxpayers covered by their large taxpayer programmes (see Figure 6.2).
Note: The table shows the average percentages across the jurisdictions that were able to provide the information.
Sources: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Tables D.17 Large taxpayer office / program ratios: Full-time equivalents (FTEs), and D.18 Large taxpayer office / program ratios: Corporate taxpayers, additional assessments raised, and net revenue administered, http://isoradata.org (accessed on 1 October 2025).
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Table D.18 Large taxpayer office / program ratios: Corporate taxpayers, additional assessments raised, and net revenue administered, http://isoradata.org (accessed on 1 October 2025).
While the management of these groups of taxpayers is often undertaken as a programme, in a large number of jurisdictions these programmes are also structural involving a Large Taxpayer Office or HNWI unit. As can be seen in Table 6.6. for large taxpayer offices / programmes, the scope of the work of these units varies considerably, ranging from undertaking traditional audit activity, through to “full service” approaches which may also encompass co-operative compliance programmes (see Chapter 8 for more on this). However, on average 60% of tax administration staff in large taxpayer offices or programmes are working on audit, investigation and other verification related issues (see Table 6.5.).
Percentage of administrations
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Tables A.34 Large taxpayer office / program: Existence and revenue collected, A.35 Large taxpayer office / program: Functions - Registration, return and payment processing, and services, and A.36 Large taxpayer office / program: Functions - Audit, debt collection, dispute resolution, http://isoradata.org (accessed on 1 October 2025).
Percentage of administrations
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Table A.40 High net wealth individuals (HNWIs) office / program: Existence and revenue collected, http://isoradata.org (accessed on 1 October 2025).
Planning for future risks
While it is key for tax administrations to understand current compliance risks and prepare appropriate response strategies, it is equally important to understand and prevent risks which may arise in the future. The increasing availability of data along with the enhanced capacity of tax administrations to handle and analyse that data allows tax administrations to more robustly assess future tax risks.
The ability to identify, understand and manage risks in a rapidly changing environment is a critical element of successful and resilient tax administration. Table 6.3. highlights the large number of tax administrations who engage in forecasting, which is putting them in a position to assess where new compliance risks may arise, and to develop appropriate mitigation strategies. This is leading to the creation of sophisticated risk management programmes, that can embed risk management across the organisation and even across government rather than being carried out in silos, as illustrated by the example from France in Box 6.6.
France have introduced a new project, Signaux Faibles (meaning weak signal), to facilitate the coordination of the General Directorate of Public Finances (Direction Générale des Finances Publiques, DGFiP) with other government agencies to detect and support companies in difficulty as early as possible, so that action can be taken to avoid bankruptcy.
This digital project is based on a partnership between five different French government agencies with an interest in supporting businesses - DGFiP, the Bank of France, the General Directorate for Business, the General Delegation for Employment, and the Social Security Agency. These five partners share their data to feed a predictive artificial intelligence algorithm by cross-referencing a range of tax and social security data. Its target is to detect companies (generally small and medium enterprises) likely to enter insolvency within eighteen months. Every quarter, the algorithm is fed the latest data, and produces a list of companies at risk of insolvency.
Signaux Faibles is accessible to approximately 1 000 authorised civil servants from the partner government agencies, and offers the ability to:
View the quarterly list of companies detected as potential for bankruptcy by the algorithm;
Consult company data; and
Exchange information relating to the support of companies in difficulty, in order to ensure everyone has the relevant information, better coordinate actors and record actions.
Once the detected companies have been further investigated, supplemented by internal data and analysis, companies are contacted proactively in order to offer assistance and suggest solutions.
Source: France (2025).
Preventing non-compliance
Tax administrations rely heavily on the positive compliance attitudes of taxpayers in reporting and paying their taxes. This is often termed “voluntary compliance”. Compliance attitudes are particularly important where tax administrations rely heavily on taxpayers to undertake full and accurate self-reporting of taxable income and taxable events and to make payments.
As highlighted in Chapter 5, tax compliance can be heavily affected by elements outside of the control of the tax administration, but they can use a variety of service-related approaches to support voluntary compliance and prevent non-compliance. Typically, those approaches take place before tax returns are filed and include:
Reminding taxpayers of deadlines (filing and paying);
Facilitating taxpayer access to third-party data already collected, for example, through pre-filing regimes or access to such data through taxpayer portals;
Running targeted campaigns to encourage compliance; and
Providing educational and support initiatives.
However, there are also non-service-related approaches that tax administrations have at their disposal to influence compliance, such as providing taxpayers with information on predetermined compliance interventions. Knowing that an intervention might come, may encourage taxpayers to pay closer attention to tax compliance issues.
Behavioural insights and nudges
Another approach for preventing non-compliance is the use of behavioural insights. Behavioural insights is an interdisciplinary field of research using principles from the behavioural sciences such as psychology, neuroscience, and behavioural economics to understand how individuals absorb, process, and react to information. These principles can be used to design practical policies and interventions based on human behaviour. This can be particularly powerful when combined with insights gathered from the analysis of the increasingly large volumes of data available to tax administration, both internally and externally generated. One example of this are nudge messages during the return filing process, providing taxpayers with an indication where there might be potential issues/ errors in the figures being reported.
As noted in Table 10.4. of the 2024 edition of this series, around half of the administrations report employing behavioural researchers (OECD, 2024[3]) and the 2021 report Behavioural Insights for Better Tax Administration: A Brief Guide prepared by the FTA’s informal Community of Interest on Behavioural Insights contains many examples of this in practice (OECD, 2021[9]). Box 6.7. looks at some of the ways in which administrations are using behavioural insights and nudges to make taxpayers aware of their obligations and increase compliance.
Canada – Taxpayer behavioural analytics
The CRA Digital Services allow Canadians to sign up for email notifications when creating a CRA My Account. Upon registration, users receive a confirmation email encouraging them to complete their My Account setup if they have not already.
To enhance My Account registrations, the CRA conducted a behavioural insights campaign in collaboration a post-secondary institution. This four-week experiment in June 2022 targeted approximately 900 000 participants, who received one of nine different follow-up email messages testing various behavioural interventions. The initiative was highly successful, resulting in over 90 000 new My Account registrations and valuable insights into user behaviour.
In March 2024, the CRA launched a new Digital Identity Validation service, enabling Canadians to register and access CRA sign-in services in real-time, thus eliminating the need to wait up to 10 business days for a security code by mail.
Building on this success, in January 2025, the CRA partnered with academics to initiate the second phase of the experiment. This phase targeted individuals who had registered for email notifications but had not completed their My Account registration. By utilizing the new Digital Identity Validation service, the CRA aimed to encourage these users to finalize their registration and access their online mail, ultimately enhancing their overall experience with CRA services.
Türkiye – VAT and Stamp Tax compliance measures
The Turkish Revenue Administration regularly sends informative SMS messages to encourage taxpayers to file their VAT returns and make both Stamp Tax and VAT payments before the deadline. This communication strategy, which is based on behavioural insights, aims to raise taxpayers' awareness and help them adopt the correct behaviours.
These messages are used as an effective tool to enhance taxpayers’ compliance levels. The messages emphasise the legal consequences, financial losses, and potential penalties that may arise if obligations are not met, thereby supporting taxpayers in making informed decisions. Additionally, the SMS messages provide practical information on how to fulfil these obligations, simplifying the process for taxpayers.
In 2024, SMS notifications were sent to a total of 7 295 513 VAT taxpayers. Among them, 1 211 782 messages were sent regarding VAT principal obligations, and 6 083 731 messages were sent concerning Stamp Tax responsibilities. SMS reminders were sent every month to taxpayers who had not made their payments within five days after the deadline. Approximately five days after the SMS is sent, collection/accrual rates are measured based on the payments made by taxpayers. The average increase in the collection/accrual rate was calculated at 13.37% for VAT, and 15.30% for Stamp Tax, demonstrating the positive effects of this approach.
Sources: Canada (2025) and Türkiye (2025).
Taxpayer rating programmes
Over the past years, there has been an increasing number of administrations reporting the introduction of taxpayer rating programmes to encourage compliance through instilling a sense of responsibility around paying taxes and providing indicators for taxpayers to measure how well they are complying with their obligations. To further incentivise compliance, this is sometimes accompanied with a rewards system for those who comply. Box 6.8. contains an update from Latvia on one such programme.
In the 2024 edition of this publication (OECD, 2024[3]), Latvia’s State Revenue Service (SRS) outlined its taxpayer ratings for companies, which enable companies to track their compliance performance and improve it. Each company can access its rating and an explanation of how it is formed in their Electronic Declaration System profile.
This system is being continuously developed with new improvements. Now, anyone interested can view a company’s taxpayer rating overall assessment in the SRS public database. More detailed information about taxpayers’ rating indicators is available only to a company in their Electronic Declaration System profile and is not publicly available. These indicators help companies to understand where and how they can improve their rating. However, a company can share this detailed information on their taxpayer’s rating with business partners and public bodies, building trust in the tax community. The SRS provides cooperation benefits for companies with higher ratings, whilst restrictions are in place for companies with a lower rating.
A detailed description of the taxpayer rating system and criteria used for evaluation is now also publicly available on the SRS website to ensure transparency. The continuous development of the Taxpayer’s Rating System includes improvements to the evaluation criteria and calculation algorithm.
Source: Latvia (2025).
As regards preventing non-compliance, it is important to note that many of the other Chapters also include information on the work tax administrations are doing to positively influence tax compliance, for example, pre-filling regimes (Chapter 4), taxpayer services (Chapter 5), or mechanisms to prevent disputes (Chapter 8). However, it is essential to be aware that this is only a glimpse of the work tax administrations are doing to ensure compliance.
Tax administrations determine through a combination of methods and tools whether a taxpayer is non-compliant with their obligations. This may include data matching programmes; data analytics and the use of algorithms, for example, as part of compliance risk models; information sharing between government agencies and jurisdictions; and risk reviews where officials might look into public records and social media.
Based on those methods and tools, a tax administration might request further information or provide a taxpayer the opportunity of voluntary disclosure. Where deemed necessary, a “compliance action” will be launched to determine whether taxpayers have properly reported their tax liability.
Electronic compliance checks
The increasing availability of data and the introduction of sophisticated analytical models and artificial intelligence are allowing administrations to better identify returns and claims or transactions which might require further review or be fraudulent. Furthermore, these models, many of which can operate in real-time, are allowing administrations to conduct automated electronic checks on all returns or on transactions of a particular type.
The use of automated electronic checks or using rules-based approaches to treat some defined risks (for example, automatically denying a claim, issuing a letter or matching a transaction) is providing administrations with more effective and efficient ways to undertake some of this work. As Table 6.8. indicates, more than 90% of administrations are using electronic compliance checks as part of the return filing process, with:
Around two-thirds of those doing this during the process of completing the return or while submitting it, for example, via prompts and real-time nudging indicating that information might be missing or deductions to high; and
More than 85% using electronic checks post return submission.
Percentage of administrations that undertake the relevant approach
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Table A.86 Verification / audit activity: Electronic compliance checks, http://isoradata.org (accessed on 1 October 2025).
Audits
On average, at around 60% audit adjustment rates have remained stable over the period 2018 to 2023 (see Table 6.9.). However, as shown in Figure 6.3., the rates vary significantly across the administrations covered by this report.
Note: The table shows the average audit adjustment rates and additional assessments raised through audits (excluding electronic compliance checks) for those jurisdictions that were able to provide the information for the years 2018 to 2023. The number of jurisdictions for which data was available is shown in parentheses.
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Table D.46 Audit ratios: Hit rate and additional assessments raised, http://isoradata.org (accessed on 1 October 2025).
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Table D.46 Audit ratios: Hit rate and additional assessments raised, http://isoradata.org (accessed on 1 October 2025).
A direct comparison of the audit adjustment rates for the 10-year period 2014 to 2023 is not possible as the 2014 ISORA data distinguished between different audit types and the adjustment rates for the total number of audits completed are only available for a few jurisdictions. However, looking at the adjustments rates by audit type, shown in Table 6.10., it appears that the 2014 figures are very similar to the overall adjustment rates reported in Table 6.9. for the period 2018 to 2023.
Source: OECD (2017), Tax Administration 2017: Comparative Information on OECD and Other Advanced and Emerging Economies, Table A.16, https://doi.org/10.1787/tax_admin-2017-en.
The importance of audits can be seen when looking at the additional assessments raised. Between 2018 and 2023, the additional assessments raised from audits were on average between 3% and 4.2% of total revenue collections. However, while this has been relatively flat over the years 2018 to 2021, it declined noticeably in 2022 and 2023 (see Table 6.9.). Looking at the jurisdiction level data, it can be seen that there are significant differences across the 54 administrations that were able to provide data (see Figure 6.4.).
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Table D.46 Audit ratios: Hit rate and additional assessments raised, http://isoradata.org (accessed on 1 October 2025).
Figure 6.5. looks at the distribution of the ratio of additional assessments raised to tax collected across jurisdictions for the main tax types. The boxes depict the 2nd and 3rd quartile (i.e. the central 50% of jurisdictions) with the middle line indicating the median value. This shows that:
The distribution of the values is the greatest for CIT with a median of around 4%;
For PIT and VAT, the values are very similar for the central 50% of jurisdictions: for PIT between 0.3% and 2.8% and for VAT between 0.6% and 2.7%;
The values are the lowest for PAYE, between 0.05% and 0.8% for the central 50% of jurisdictions;
There are some significant outliers, and it should be noted that for CIT the figure does not show two outliers that are above 20%.
In many jurisdictions, the additional assessments raised through LTO/Ps make-up a significant share of the total additional assessments raised from audits (see Figure 6.6.). On average, LTO/Ps contribute around one-third of the total additional assessments raised from audits (see Table 6.5.).
Note: As regards CIT, the figure does not show the outliers above 20% even though those data points were used for the underlying calculations.
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Tables D.47 Audit ratios: Additional assessments raised by tax type - CIT and PIT, and D.48 Audit ratios: Additional assessments raised by tax type - PAYE and VAT, http://isoradata.org (accessed on 1 October 2025).
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Table D.18 Large taxpayer office / program ratios: Corporate taxpayers, additional assessments raised, and net revenue administered, http://isoradata.org (accessed on 1 October 2025).
The ISORA data provides insight into the change in the additional assessments raised from audits by tax type for the 10-year period from 2014 to 2023. As can be seen in Table 6.11., the average values decreased significantly for PIT, CIT and VAT, which could hint at effective up-front compliance programmes and improved compliance behaviour, potentially a result of better services and taxpayer education programmes, and a deterrent effect with taxpayers understanding that tax administrations hold an increasing amount of data.
Note: The table shows the average of additional assessments raised through audit as percentage of tax collected for those jurisdictions that were able to provide the information for the years 2014 and 2023. The number of jurisdictions for which data was available is shown in parentheses.
Sources: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Tables D.47 Audit ratios: Additional assessments raised by tax type - CIT and PIT, and D.48 Audit ratios: Additional assessments raised by tax type - PAYE and VAT, http://isoradata.org (accessed on 1 October 2025), and calculations based on OECD (2017), Tax Administration 2017: Comparative Information on OECD and Other Advanced and Emerging Economies, Tables A.17, A.28, A.29, A.161 and A.162, https://doi.org/10.1787/tax_admin-2017-en.
Tax crime investigations
As mentioned in Chapter 2, close to two-thirds of the tax administrations covered in this publication are involved in conducting tax crime investigations, and the ISORA survey asked them to provide information regarding the number of cases referred for prosecution.
Table 6.12. shows the total number of cases referred for prosecution during the fiscal year for the 30 administrations that have responsibility for conducting tax crime investigations and that were able to provide the data for the years 2018 to 2023. The annual reduction of the total number of cases referred for prosecution that is visible between 2018 and 2022 has come to an end in 2023 when the total number of cases referred for prosecution increased by 15% compared to 2022. However, the 2023 value is still well below the values reported for 2018 to 2021.
Note: Only includes data for administrations that have responsibility for tax crime investigation and were able to provide the information for the years 2018 to 2022.
Source: ADB, CIAT, IOTA, IMF, OECD, International Survey on Revenue Administration, Table A.88 Tax crime investigations: Number of cases, http://isoradata.org (accessed on 1 October 2025).
This is also reflected in the jurisdiction level data, which shows that 60% of administrations with responsibility for conducting tax crime investigations reported an increase in the number of cases referred for prosecution between 2022 and 2023 (see Table A.88).
Looking at the 10-year period between 2014 and 2023, data for 23 jurisdictions shows that in almost three-quarters of administrations the number of cases referred for prosecution decreased. (Calculation based on Table A.135 in the 2017 edition (OECD, 2017[10]) and Table A.88 in the ISORA database.)
References
[3] OECD (2024), Tax Administration 2024: Comparative Information on OECD and other Advanced and Emerging Economies, OECD Publishing, Paris, https://doi.org/10.1787/2d5fba9c-en.
[4] OECD (2023), International Standards for Automatic Exchange of Information in Tax Matters: Crypto-Asset Reporting Framework and 2023 update to the Common Reporting Standard, OECD Publishing, Paris, https://doi.org/10.1787/896d79d1-en.
[6] OECD (2023), Tax Administration 2023: Comparative Information on OECD and other Advanced and Emerging Economies, OECD Publishing, Paris, https://doi.org/10.1787/900b6382-en.
[8] OECD (2022), Analytics Maturity Model, OECD, Paris, https://www.oecd.org/en/topics/sub-issues/comparative-analysis-of-tax-administrations/tax-maturity-models.html (accessed on 1 October 2025).
[9] OECD (2021), Behavioural Insights for Better Tax Administration: A Brief Guide, OECD Publishing, Paris, https://doi.org/10.1787/582c283e-en.
[10] OECD (2017), Tax Administration 2017: Comparative Information on OECD and Other Advanced and Emerging Economies, OECD Publishing, Paris, https://doi.org/10.1787/tax_admin-2017-en.
[2] OECD (2017), The Changing Tax Compliance Environment and the Role of Audit, OECD Publishing, Paris, https://doi.org/10.1787/9789264282186-en.
[7] OECD (2016), Advanced Analytics for Better Tax Administration: Putting Data to Work, OECD Publishing, Paris, https://doi.org/10.1787/9789264256453-en.
[5] OECD (2015), Transfer Pricing Documentation and Country-by-Country Reporting, Action 13 - 2015 Final Report, OECD/G20 Base Erosion and Profit Shifting Project, OECD Publishing, Paris, https://doi.org/10.1787/9789264241480-en.
[1] OECD (2004), Compliance Risk Management: Managing and Improving Tax Compliance, OECD, Paris, https://www.oecd.org/content/dam/oecd/en/topics/policy-issues/tax-administration/compliance-risk-management-managing-and-improving-tax-compliance.pdf (accessed on 1 October 2025).