Countries are strengthening their anti-fraud functions at the strategic, institutional and operational level to support existing integrity safeguards in the face of increasing fraudulent activity. While many governments have basic safeguards in place, such as embedding anti-fraud and corruption in internal control and risk management, most countries lack a systemic approach for improving the coherence, targeting and effectiveness of their anti-fraud efforts. There is increasing recognition of the potential of emerging technologies, such as AI, to fight increasingly complex and transnational fraud schemes. Overall, as governments seek public sector efficiencies, investing in fraud prevention will be more cost-effective than dealing with its consequences.
Anti‑Corruption and Integrity Outlook 2026
Abstract
Governments are facing growing pressure to restore public finances while increasing spending on strategic priorities amid tight fiscal constraints and rising levels of global public debt. (OECD, 2025[45]) Safeguarding public resources from waste and abuse, including fraud, has therefore never been more important. However, resources for protecting public funds are often not ringfenced and anti-fraud functions are asked to “do more with less” in periods of fiscal austerity. Paradoxically, even when presenting a high return on investment, governments may scale down anti-fraud spending.
Measurement of fraud losses and savings gained from preventive efforts are rapidly improving. Still, most countries do not have reliable data on the share of public funds that are lost to errors, irregularities and fraud instead of improving public services. In those countries that collect such data, a clear trend is emerging: the scale and nature of fraud are rapidly evolving. The Association of Certified Fraud Examiners’ (ACFE) latest Global Fraud Survey showcased that organisations are estimated to lose 5% of funds globally to occupational fraud each year (over USD 5 trillion in losses) (Association of Certified Fraud Examiners (ACFE), 2024[6]) In several countries, fraud has become the most prevalent and fastest growing type of crime in recent years:
In New Zealand, fraud is amounted to 29% of all crimes committed, according to data from the New Zealand Crime and Victims Survey in 2024
According to the Swedish National Council for Crime Prevention, fraud is the category of crime that increased the most in 2023. (New Zealand Government, 2023[46]); (Sveriges Riksbank, 2024[47])
Fraudsters are using technology and transnational networks to increase the scale and sophistication of their methods. To effectively respond to this threat, governments are increasingly shifting the dial from agency-specific interventions to a systematic, whole-of-government approach to counter fraud, where resources and expertise are co-ordinated across different agencies and levels of government. Furthermore, investing in capacity-building, digitalisation and the use of data analytics, including artificial intelligence (AI) to address fraud, as well as ensuring the interoperability of systems and cross-agency data sharing, is imperative for timely fraud prevention, detection, and response. There is a growing recognition that, given the speed and complexity of modern-day fraud, investing in preventive tools to stop the fraud before it occurs is more cost-effective than the traditional “pay-and-chase” approach, which focuses primarily on investigating individual cases of large-scale fraud. This focus chapter finds that:
The funds lost to public sector fraud are substantial, and increasing
While governments have basic safeguards in place, most countries lack a strategic framework to address fraud
Investments in emerging technologies and AI are key to counter increasingly complex and transnational fraud schemes
While governments often rely on traditional enforcement methods, there is an increasing understanding that investing in fraud prevention is more cost-effective than dealing with the consequences once fraud has occurred.
What is fraud?
There is not a single, universally accepted legal definition of “fraud” that applies globally, with countries’ criminal laws describing fraud in different ways and to varying degrees of specificity. However, there are certain core elements in defining fraud across jurisdictions, namely that fraud uses deception (knowing misrepresentation of the truth or concealment of a fact) to achieve a financial or other material gain, and that it causes a detriment to a person or organisation (UNODC, 2024[48]); (OECD, 2020[49]).
Fraud can be:
Internal – the use of one’s occupation for personal enrichment through the deliberate misuse or misapplication of the employing organisation’s resources or assets. Fraud may, but does not always, include abuse of public office; fraud is characterised by intentional deception or misrepresentation intended to secure financial or personal gain, or to cause loss to another party.
External – carried out by individuals or entities outside an organisation, with the intent to deceive or exploit the system for financial or other illicit gains.
Targeted towards an individual – such as through identity theft or phishing scams.
Targeted towards an organisation – such as embezzlement or financial statement fraud when committed internally, or social welfare fraud, inflated invoices from external vendors, or theft of intellectual property or customer information when committed externally (ACFE, 2025[50]).
Fraud in the public sector could involve an individual or entity using false information with the intention of obtaining government funds, information, goods or services to which they are not entitled, or to use for other than the intended purpose. For example, social benefit programmes are typically vulnerable to fraud due to the high volume of transactions, complex eligibility criteria and reliance of self-reported information which can be difficult to verify, such as income or employment status. This creates opportunities for fraudsters to exploit loopholes, misrepresent their circumstances, or fabricate documents to receive benefits they are not entitled to (OECD, 2020[49]).
The Association of Chartered Certified Accountants (ACCA) has conducted extensive research on the prevalence and materiality of fraud in the public sector, notably as part of a recently conducted global survey from over 30 roundtable discussions with risk and finance professionals across sectors and regions. According to the study “Combatting Fraud in a Perfect Storm”, procurement fraud, abuse of authority fraud, third-party fraud, expense fraud and cyberfraud are some of the most prevalent types of fraud in the public sector. In parallel, procurement fraud, other insider fraud, bribery & corruption, third-party fraud and expense fraud were the most material (i.e. have the largest impact) on the public sector. (Association of Chartered Certified Accountants (ACCA), 2025[51]).
Furthermore, growing risks of serious and organised crime to commit widespread, systematic fraud poses further threats to governments, not least by exploiting government programmes and procurement systems, but also in the form of co-ordinated scams or targeted cyber-attacks. According to Europol, fraud constitutes the most rapidly expanding sector in organised crime, targeting a broad spectrum of victims, including individuals, public and private sector organisations, and their data (Europol, 2025[52]).
Benefit and welfare fraud
This could involve deliberately misrepresenting circumstances or failing to disclose a change in circumstances to qualify for benefits the claimant is not entitled to (e.g. undeclared income or partner) or applying for multiple welfare benefits from different programmes without disclosing that similar benefits are received elsewhere (double dipping).
Grant fraud
When individuals or entities knowingly misrepresent or fail to disclose information to obtain or misuse public funds awarded through grants.
Synthetic identity fraud
Fraudsters, including organised criminal networks, can create synthetic identities using a combination of real and falsified data to gain access to public services (e.g. collecting multiple benefit cheques or submitting fraudulent health care claims), including with the help of AI.
Third-party provider fraud
Service providers could submit fraudulent invoices or claims for services that were never rendered, or for goods and services that do not meet program standards.
Cyber fraud
When individuals or entities target public organisations through co-ordinated scams or attacks to obtain sensitive information for profit, for example intellectual property or customer information.
Tax fraud, including cross-border VAT fraud
Where an individual or entity deliberately misrepresents income or falsifies documents to avoid paying tax or pay less tax than required. This could range from an individual committing tax evasion to multiple entities, including organised criminal groups, engaging in e.g. VAT carousel fraud (where goods are acquired free of VAT and resold on the domestic market inclusive of VAT, and the VAT is not paid to the national authorities).
Procurement fraud
When individuals or entities manipulate the procurement process or procure low quality items, issue false invoices or receive kickbacks for referring contract work to related parties.
Double Funding
When a single cost for a single activity is publicly funded by multiple instruments, such as an entity receiving public funding from more than one agency to purchase a single item.
Fraudulent expenditure claims
When individuals of a public sector organisation use false receipts to seek reimbursement for a business expense that was not incurred or inflates the actual cost of the expense.
Measuring the true financial scale of fraud remains a challenge in many countries, but tools are improving to estimate losses due to fraud. There are now credible examples from the public sector to estimate the financial impact of fraud in government schemes, which indicate that the actual cost of fraud is much higher than the levels of detected fraud (Box 9.2). This provides a powerful impetus for governments to reinforce measures to counter fraud. Yet, the cost of fraud is not just financial. Fraud diverts taxpayers’ money away from essential government services, resulting in fewer, substandard, or less safe services delivered to those who need it the most. Ultimately, fraud in the public sector can erode citizens’ trust in government institutions, create opportunities for further exploitation, decrease legal compliance and increase the public’s tolerance for fraud (International Public Sector Fraud Forum, 2020[58]) (Department for Work & Pensions, 2023[59]).
UK Cabinet Office, National Audit Office and Public Accounts Committee
According to a Cabinet Office review of around 50 fraud and error measurements, public bodies are estimated to lose between 0.5 and 5% of their spending to fraud and related loss, which amounts to GBP 31 billion to 48 billion annually (approximately USD 42.3 billion to 65.5 billion annually). The UK National Audit Office conducted an overview of the impact of fraud and error on public funds, where it estimated that fraud and error cost taxpayers between GBP 55 billion and 81 billion (approximately USD 75 billion and 111 billion) in 2023-2024.
The COVID-19 Bounce Back Loan Scheme involved business entities fraudulently abusing government subsidies in the form of loans that they were not eligible for, amounting to an estimated loss of GBP 4.9 billion (approximately USD 6.6 billion, over 10% of the loans provided).
US Government Accountability Office and Department of Justice
The US Government Accountability Office (GAO) estimated that the amount of fraud in unemployment insurance (UI) programs during the COVID-19 pandemic was likely between USD 100 billion and USD 135 billion (11% and 15% of the total amount of unemployment insurance benefits paid during the pandemic).
The Department of Justice’s National Health Care Fraud Takedown in 2025 showed USD 14.2 billion in intended loss from several health care fraud schemes, including the involvement of transnational criminal organisations which were alleged to have submitted over USD 12 billion on fraudulent claims to Medicare, the United States’ health insurance program.
Australia’s Institute of Criminology
According to a census prepared by the Australian Institute of Criminology, where responses were collected from 157 Australian Government entities, Commonwealth entities reported external fraud losses of approximately AUD 105 million (approximately USD 67 million) in 2023-2024. However, the estimated financial loss to fraud in Australia is AUD 5 billion to 25 billion (approximately USD 3.2 billion to USD 16 billion) per year.
European Public Prosecutor’s Office
The 2025 annual report of the European Public Prosecutor’s Office (EPPO) stated that, at the end of 2024, the damage resulting from fraud and other crimes against the EU budget was estimated to be approximately EUR 25 billion (an increase of 22.5% compared to the previous year). More than half of the estimated damage was linked to cross-border VAT fraud, which included the systematic involvement of criminal organisations.
Occupational fraud in government organisations
The ACFE carries out the largest global study on occupational fraud each year. The global study includes data from 138 countries and territories and 22 major industry categories from public, private and non-for-profit organisations around the globe. According to its latest report (published in 2024), the median loss due to fraud in government organisations was approximately USD 150 000, whereas the average loss was approximately USD 2.3 million. Median losses were largest at the national level of government.
Source: (National Audit Office, 2024[60]); (UK Parliament, 2022[61]) (United States Government Accountability Office, 2023[62]); (U.S. Department of Justice - Office of Public Affairs, 2025[63]) (Merran McAlister and Samantha Bricknell, 2025[64]); (International Public Sector Fraud Forum, 2020[58]); (European Public Prosecutor’s Office, 2025[65]); (Association of Certified Fraud Examiners (ACFE), 2024[6])
The rapidly evolving nature of fraud is exacerbated by emerging risks such as rapid developments in technology, where artificial intelligence is increasingly used by individuals and criminal organisations to commit fraud. Furthermore, fraud can no longer be regarded as a purely domestic issue, as shown by the rising levels of cross-border VAT fraud in the European Union. According to the 2025 annual report from EPPO, fraud and other crimes against the EU budget have increased by 22.5% compared to the previous year, with more than half of the estimated damage linked to the systematic perpetration of cross-border VAT fraud by criminal organisations (European Public Prosecutor’s Office, 2025[65]); (OECD, 2024[66]).
Governments are increasingly adopting dedicated frameworks to fight fraud, although strategic approaches vary amongst countries. Currently, 33% of OECD Member and partner countries with a national anti-corruption and integrity strategic framework have a dedicated focus on countering fraud.1 A growing number of countries are also choosing to adopt stand-alone national, sectoral and organisational anti-fraud strategies. In the European Union, the adoption of national anti-fraud strategies is partially driven by Regulation (EU) 2021/1060, which stipulates that Member States must ensure the legality and regularity of expenditures related to EU funds, and to take all necessary measures to prevent, detect, correct and report any irregularities, including fraud. The Regulation is one example of several EU requirements which require (or strongly encourage) member states to adopt effective anti-fraud measures to protect EU funds. While EU countries maintain the flexibility to cover national funds within their anti-fraud strategy, the majority of them only cover a small share (which are typically related to EU funds). The EU continues to promote a whole of government approach by recommending that all EU member states adopt a national anti-fraud strategy. This will ensure the relevant national stakeholders take a shared approach and coordinate cooperation with the EU. (European Commission, 2025[67])
Out of 63 OECD Member and partner countries, seven countries (11%) have national, comprehensive anti-fraud strategies in force, and nineteen countries (30%) have strategies adopted at the level of an individual ministry or agency (labelled “organisational-level strategies”). Another ten countries (16%) have anti-fraud strategies for a specific sector of the economy (e.g. Agriculture), an area of public spending (e.g. welfare benefits), or a dedicated funding scheme (e.g. the EU RRP funds). Thirty-four countries (54%) have neither of these types of anti-fraud strategies in place (Figure 9.1).2
Adopting a strategic approach to counter fraud can offer significant benefits, not least by allowing governments to co-ordinate resources and expertise across multiple levels of government to mitigate wasteful spending and duplication of efforts across government departments. Prioritising efforts and resources through a risk-informed approach also empowers governments to act preventatively and allocate resources to where it matters the most, while effective monitoring and evaluation mechanisms can help governments determine whether approaches to fight fraud are cost-effective. Ultimately, adopting a strategic framework to counter fraud signals a high-level commitment to the public, which can build trust in institutions and increase accountability. Given the rising and rapidly evolving risk of fraud – including transnational fraud – investing in strategic frameworks and international co-operation will be key to counter fraud effectively with a focus on prevention.
Note: The categories of anti-fraud strategies are not mutually exclusive, and countries can have several types of strategies at once. However, for visualisation purposes, only one type of strategy is displayed per country, giving preference to national-level strategies. For example, a country that has a standalone national anti-fraud strategy and anti-fraud objective(s) found within the anti-corruption strategic framework is only displayed once, for the former category.
Organisational-level strategies tackle fraud in the activities of a public agency. Sectoral strategies refer to strategies tackling a sector of the economy or an area of public spending at the national level. No up to date (as of 2025), publicly available standalone anti-fraud strategies could be found for the following OECD Member countries: AUT, CAN, CHE, DEU, EST, ISL, ISR, JPN, KOR, LUX, POL, TUR; and OECD partner countries: ARG, ARM, BIH, BOL, GTM, HND, IDN, JOR, KAZ, PER, PRY, SYC, THA, UKR, XKV*, ZMB.
Source: OECD Public Integrity Indicators database (as of 10 March 2026) and research conducted by the OECD Secretariat.
The 2024 Anti-Corruption and Integrity Outlook showed that countries generally have strong internal control, corruption risk management regulations in place, with room for improvement for those on internal audit. These regulations address the basics of fraud prevention. 72% of countries had guidelines in place on fraud and corruption prevention as part of their internal control systems. 73% of countries explicitly addressed fraud and corruption risks in their risk management frameworks. However, this area has one of the largest implementation gaps, meaning that high-quality regulations do not always get implemented according to plan. For example, only seven countries could document a consistent use of risk assessments in practice in all ministries and large central government agencies. And only eight countries conducted risk assessments with a focus on integrity risks in at least half of line ministries and large central government agencies. Only Lithuania, Latvia, Poland, Portugal, and Ukraine fulfilled both of these PII criteria (Figure 9.2).
Note: Data is taken from the following criteria: "Guidelines on fraud and corruption prevention are available and part of the IC system", "All sample organisations have conducted at least one risk assessment exercise in the past 3 years", “Roles and responsibilities for risk management and for managing integrity risks have been assigned in all budget organisations, in line with the regulatory framework", "All sample organisations have established a system for documenting the results of risk assessments, including as a minimum creating risk profiles or risk registers', and "Risk assessments for at least half of sample organisations identify integrity risks". The following countries did not answer this portion of the questionnaire: CAN, CRI, ESP, JPN, KOR, LUX, MEX, NLD, USA, PER
Data not provided: BEL, COL, DEU, FRA, GBR, HUN, ISL, ISR, ITA, NZL
Source: OECD Public Integrity Indicators database (as of 20 October 2025).
While countries are increasingly recognising the potential of AI to fight fraud, the adoption of emerging technologies and AI among integrity actors is generally low, with many AI initiatives still in the exploratory or pilot phase. (European Commission, 2025[68]) In a study conducted by the OECD in 2024 on the stage of generative AI and Large Language Models (LLM) used by integrity actors (such as anti-corruption agencies, supreme audit institutions (SAIs), internal audit and other oversight bodies), approximately 50% (30 out of 59 organisations from 39 countries) of the surveyed organisations reported that they did not use generative AI in their operations, but were exploring potential use cases. (Ugale and Hall, 2024[69]) While there is an increase in the number of institutions engaging with AI-supported tools, these remain mostly pilot projects (Figure 9.3).
Which of the following options best describes the maturity of your institution’s use of Gen AI and LLMs specifically, as a sub-domain of AI?
Source: (Ugale and Hall, 2024[69]).
The OECD’s report on ‘Governing with Artificial Intelligence’ presents further insights into the state of play and way forward in core government functions’ use of AI. The report builds on the analysis of over 200 AI use cases, of which 57% support automated, streamlined or tailored processes and services. While the roles of integrity actors such as SAIs or internal audit bodies typically cover both fraud and corruption risks in their areas of work, one of the main potential applications of AI for integrity actors lies in detecting fraudulent activities, as AI algorithms excel at applying statistical techniques to identify outliers, patterns, transactions and behaviours that deviate from established norms and that warrant further human investigation (OECD, 2025[70]) (Figure 9.4).
% of use cases per benefit, collected in 2025
Note: The benefits in this figure are not mutually exclusive (that is, one use case can yield more than one type of benefit). Thus, the sum of potential benefits observed is greater than the total number of use cases.
Source: (OECD, 2025[70]).
For instance, the OECD Inventory of Tax Technology Initiatives (ITTI) survey illustrated that the main area of application of AI in tax administrations of OECD Member countries was for the detection of tax evasion and fraud. In this field, AI is often used to detect hidden patterns of behaviour or new connections between transactions, assets or taxpayers within the data sources held by tax administrations, but is increasingly also applied to unstructured sets of data (such as handwritten documents) to detect tax evasion or non-compliance (OECD, 2025[70]) (Figure 9.5).
How is AI being used in tax administration?
Note: 29 of the 38 OECD Members report using AI in tax administration in the 2024 Inventory of Tax Technology Initiatives.
Source: (OECD, 2025[70]); OECD Data Explorer - Inventory of Tax Technology Initiatives 2024 (https://oe.cd/dx/ITTI2024).
Yet, significant implementation challenges to successfully scale up AI initiatives in public integrity and oversight remain. Some of these challenges relate to shortages in relevant skills and experience, including not only technical skills for its safe application but also capabilities in data governance, change management and leadership. Another constraint is the lack of high-quality data and restricted access to relevant information. In particular, the ability to cross-reference and analyse data is often hindered by fragmented or non-interoperable data systems, oftentimes compounded by regulatory barriers related to data protection, privacy or security concerns. Finally, a key challenge to the wider adoption of AI in integrity institutions is the absence of robust impact measurement frameworks, which makes it difficult to demonstrate the return on investment of AI solutions to prioritise further investment in such tools (OECD, 2025[70]).
Investments in emerging technologies, including AI, are key to diagnose and target increasingly complex and transnational fraud schemes. Moving forward, effective use of AI in government contexts relies on robust transparency and accountability mechanisms to ensure that AI-driven decisions are explainable, traceable and appropriately documented. Approaches could involve developing platforms for exchanging best practices in the use and adoption of AI across institutions, providing actionable guidance and frameworks for implementing AI, and ensuring that data operations can be streamlined for enhanced communication and integration between systems. In addition, systematic documentation of outputs and reporting on the impact of AI supports the ongoing refinement of AI systems and approaches, while helping to demonstrate the value of further investment in such technologies to enhance counter fraud efforts (OECD, 2025[70]).
Fraud Risk Assessment AI Accelerator
The UK Public Sector Fraud Authority in the United Kingdom has developed an AI powered tool that allows public officials to generate draft Fraud Risk Assessments (FRAs) using Large Language Models (LLM). The tool outputs fraud risks by actor, action, and outcome based on the input material uploaded, such as guidance documents or government grant scheme briefs. This allows qualified fraud risk assessors to use the application to develop an initial FRA.
Detecting fraudulent activities and irregularities with AI
The OECD’s ‘Governing with Artificial Intelligence’ report highlights several initiatives using AI to identify anomalies and behaviours that deviate from established norms, including:
A proof-of-concept developed by the OECD and Spain’s General Comptroller (IGAE) using advanced analytics and machine learning to detect corruption or fraud risks.
The use of Machine Learning Algorithms by the European Court of Auditors (ECA) to check the transparency register of the European Commission and identify outliers as part of an audit on EU lobbying activities, with the purpose of isolating a risk-based sample for analysis by auditors
Brazil’s use of a tool to analyse daily purchasing and procurement processes to uncover risk areas and inconsistencies, where unusual patterns trigger an alert that suspends the purchase and enables further inquiry.
The use of AI to identify patterns in claims that could suggest fraud and error in the United Kingdom’s Department of Work and Pensions, so that the claims can subsequently be reviewed by relevant teams within the department.
The use of data analytics and machine learning by Portugal’s Tribunal De Contas
With support from the OECD and NOVA University Lisbon, the Tribunal de Contas developed and refined a risk assessment methodology, including the development of a data-driven risk model to undertake audit assessments. The initiative aims to improve the Tribunal de Contas’ identification of risks and the early detection of irregularities through advanced data analysis and machine learning. The methodology developed marks a significant milestone in the Tribunal de Contas’ digital transformation, refines its audit selection process and increases the effectiveness and efficiency of the public procurement system.
It is difficult to make the case for investing in reducing a problem that is not measured. Prioritising limited resources towards investments in counter fraud tools such as anti-fraud strategies, data analytics and AI requires governments to showcase the return on investment (ROI) in adopting such measures. Yet, many countries lack internally recognised methodologies to measure fraud rates and consequently to establish a baseline to showcase the ROI on counter fraud tools. As a result, reactive responses such as “pay-and-chase” approaches, where detected fraud is investigated and attempts to recover funds lost to fraud are initiated, may be more accepted due to the tangible savings generated by investigations. Establishing a well-defined measurement and evaluation framework to assess the contributions of counter fraud efforts, including AI, is therefore essential to determine whether investments in counter fraud efforts are cost-effective and subsequently enable further investments in such efforts.
Some countries are undertaking efforts to demonstrate a higher ROI through fraud prevention measures. Estimating the financial impact of fraud establishes a baseline which can help inform whether approaches to tackling fraud are cost-effective. For example, some public agencies report fraud estimates (attempting to put a value on the total extent of fraud in an area) using statistical sampling, modelling, and benchmarking to demonstrate the value of fraud prevention efforts, whereas others set annual targets for fraud reduction or prevention measures to facilitate the measurement of fraud. In addition, evaluating the non-financial impact of anti-fraud efforts and frameworks can support risk owners to determine the effectiveness of existing measures and adapt as needed. The United States Government Accountability Office recently published a technical appendix setting out approaches to evaluate effectiveness of fraud risk management in federal programs, which includes practical examples of evaluation approaches for each component of the fraud risk management framework, including on how to perform ROI calculations and measure cost savings from fraud prevention (UK National Audit Office, 2025[73]); (OECD, 2020[49]); (U.S. Government Accountability Office, 2026[74]).
Governments’ efforts to showcase the ROI of fraud prevention indicates an increasing understanding that investing in fraud prevention is more cost-effective than dealing with its consequences. In the United Kingdom, a new failure to prevent fraud offence came into force as of 1 September 2025. The new offence is a part of a wider government ambition to reduce fraud and encourage organisations to build an anti-fraud culture and move towards implementing prevention procedures, similar to the failure to prevent bribery legislation introduced in 2010. Under the new offence, large organisations (public and private) incorporated or formed by any means in the United Kingdom, or bodies incorporated and partnerships formed outside the United Kingdom but with a UK nexus, may be held criminally liable where an employee, agent, subsidiary, or other “associated person” commits a fraud with the intention to benefit the organisation or its clients and the organisation did not have reasonable fraud prevention procedures in place. Relevant organisations will have defence if they can demonstrate that they have reasonable procedures in place to prevent fraud, or that it was not reasonable in all the circumstances to expect the organisation to have any prevention procedures in place (Home Office, 2025[75]).
Similarly, Portugal introduced a law in 2021, which came into force in 2024, obliging large and mid-size companies to take preventive actions on corruption and related offences (Regime Geral de Prevenção da Corrupção, RGPC), the latter of which includes fraud-adjacent misconduct such as embezzlement or subsidy fraud. Preventive actions include conducting a risk assessment and creating and implementing a prevention plan (Presidency of the Council of Ministers, 2021[76]).
International Public Sector Fraud Authority (IPSFF) Framework on Fraud Loss Measurement
The IPSFF has developed a framework setting out key principles and processes for conducting Fraud Loss Measurement (FLM) exercises within public sector organisations. The framework:
Broadly defines the objectives of Fraud Measurement and specifically defines the objectives and purpose of FLM exercises.
Describes the steps required to plan a FLM exercise.
Explains the statistical sampling knowledge and the techniques required to undertake a FLM exercise.
Explains how to identify and describe how to use evidence to test for fraud.
Describes how estimation and measurement are used in FLM exercises.
Describes the importance of stakeholder engagement and reporting in FLM exercises.
The framework is intended to be used by counter fraud functions across all public sector organisations, with a view to support counter fraud professionals to conduct FLM exercises that produce credible estimates of the levels of fraud and error related to a specific government program, activity or function.
Forthcoming IPSFF Framework on Fraud Prevention Savings
The IPSFF is currently developing a Fraud Prevention Savings Framework, which sets out key principles and credible approaches to estimating and quantifying the financial and non-financial benefits of fraud prevention and compliance activities, as well as how to validate and communicate the results. The framework:
Outlines how fraud prevention is a key part of achieving the optimal cost-effective level of fraud control.
Sets out different types of preliminary activities that can help with measuring fraud prevention savings.
Identifies the types of savings that can be attributed to different types of fraud prevention and compliance activities.
Provides models and approaches for:
a. developing counterfactuals
b. estimating the effect of an intervention
c. quantifying the resultant savings
Provides advice on how to communicate results to key decision makers.
Outlines governance arrangements to provide a level of assurance that approaches are robust and consistent.
Explains when and why to undertake a retrospective measurement and review.
← 1. This percentage refers to countries that include objectives with an explicit anti-fraud focus in their national anti-corruption and integrity strategic framework (15 out of 46 countries).
← 2. This data is based on up to date (as of 2025), publicly available standalone anti-fraud strategies at the national level.