Fiscal Policy / WORLD

OECD Finds Digital Tax Services Expanding Across 108 Economies

The OECD’s new review compares digital identity, data sharing and artificial intelligence across 108 economies. It says the findings come from a 2024 survey and rely on information reported by administrations themselves.

Reading sizeSources ↓Share story ↗
Conceptual illustration of public finance and resource allocation.Illustration
Symbolic illustration

The OECD published a 112 page report on 12 November examining digitalisation in tax administrations across 108 economies. It finds that digital identity, information exchange between public bodies and online services are established in many places. Artificial intelligence use is more limited, and safeguards are not equally common. The central message is that transformation involves more than buying software. Identity verification, data access, work processes, staff capabilities and governance have to change together. This publication is neither a census of every tax authority in the world nor an independent audit of how effectively digital systems work.

The scope matters when interpreting the figures. The OECD Inventory of Tax Technology Initiatives draws on responses to the Global Survey on Digitalisation, launched in 2024 and completed by 108 economies. The OECD explicitly says that neither it nor its partner organisations reviewed or validated the underlying data, which should be treated as information reported by the economies themselves. Although the report appeared in November 2026, collection began earlier and many tables label indicators as relating to 2025. The findings therefore are not evidence of a fresh survey wave. They are a snapshot that must be read with the reporting dates and coverage in view.

The OECD compares the economies through seven regional groups. They include 17 in Africa, 10 in Central and West Asia, 20 in East and South Asia, 14 in Central and East Europe, 19 in North, South and West Europe, 16 in Latin America and the Caribbean, and eight Pacific Island economies. The groups add up to 104. Australia, Canada, New Zealand and the United States are in the inventory but are not assigned to a region, although they enter overall averages. Regional percentages therefore rest on groups of eight to 20 economies. These differences should not be presented as a fixed country ranking or as proof that every country in a region has the same capacity.

The change in coverage also affects historical comparisons. An OECD digitalisation report published in June 2025 focused on 54 members of the Forum on Tax Administration. The new report covers 108 economies and includes every participant in that forum. The broader scope allows comparison across more jurisdictions, but placing samples of different sizes next to each other does not by itself show progress or decline. That would require repeated survey waves covering the same administrations with the same definitions and reference dates. The OECD presents this publication as a regional overview and a resource for policy learning, not as an impact evaluation of how much each country has changed over time.

Digital identity is one of the most widespread administrative requirements in the report. The OECD says an average of 86% of administrations require individuals to use an approved digital identity to reach secure services. The share for businesses is 82%. These are shares of administrations reporting a requirement, not the percentage of taxpayers who use a digital identity. Identity verification can help establish who is filing, paying or sharing information with a public body. But having an identity system is different from connecting it across services. The report shows that such links are not equally common, and that legal and administrative choices matter alongside technical infrastructure.

Application programming interfaces, or APIs, are another way to connect tax services to software used by businesses. More than 80% of administrations in the inventory report developing APIs, and over half make them available to third parties. The OECD says every participating administration in Africa and Central and West Asia is developing APIs, while public access is more common in Europe. Development alone does not show that businesses can use an API, that systems follow compatible standards or that reporting burdens have fallen. About three quarters of administrations measure taxpayer experience, a reminder that service performance cannot be judged by counting technical links alone.

Data flows form another layer of digital tax administration. Around 60% of administrations say they receive data directly from taxpayer business systems beyond pay as you earn reporting. A greater share receive data from third parties. More than 80% report exchanging bulk information with other government agencies, and over half of the economies operate shared databases across government. Such arrangements can reduce repeated requests for the same information, but results depend on data quality, legal authority, secure access and a way to correct inaccurate records. The report measures whether exchanges exist. It does not establish how much time taxpayers save or whether error rates have changed.

Pre filled returns offer a visible example of how data integration can affect taxpayers. Many administrations can fully pre fill personal income tax returns for some parts of the taxpayer population. Similar arrangements for value added tax and corporate income tax are less common, although electronic invoicing systems support progress. Latin America and the Caribbean is identified as the region with the highest use of pre filled business tax returns. Taxpayers still need to review and correct information, and rules have to be understandable. Automation does not guarantee an accurate obligation. The records used to pre fill a return must also be reliable.

The artificial intelligence findings highlight a gap between adoption and governance. Forty nine percent of administrations report using AI to strengthen compliance or improve services. Fraud detection and risk assessment are prominent applications. Among administrations that report AI use, 85% say they have put limitations in place, while 40% report an ethical framework. These two figures refer to the AI using group, not all 108 economies. On the reported data, limitations are more common and formal ethical frameworks less common. The report also describes internal review processes. Self reported information, however, cannot independently demonstrate that systems are unbiased or that appeal rights work in practice.

Whether digital transformation lasts depends on implementation capacity as well as tools. About three quarters of administrations report a digital transformation strategy, more than 70% have a senior governance structure and about two thirds have dedicated digital leadership roles. Yet only 58% say they have funding set aside for the strategy period. That gap points to a difference between a policy document and the resources needed to deliver it. Staff skills, project delivery and security capacity also matter. The OECD says comprehensive workforce planning remains limited. A strategy does not guarantee investment, skilled staff or public accountability.

Regional examples require careful definitions. One of the eight Pacific Island economies reported requiring digital identity for individuals to access secure tax services, equal to 13%. This does not mean that digital identity is never used elsewhere in the region. The indicator counts whether a particular administrative requirement exists. Comparisons of APIs, data exchange or AI also have to account for sample sizes and differences in law. The OECD cautions that tax systems operate within different policy and legislative environments, and a single administrative approach may not suit every jurisdiction.

The report is best read as a starting map rather than a scorecard of results. Wider use of digital identity, links between institutions and online services may provide foundations for more integrated tax administration. The publication does not show how much these technologies increase revenue, reduce compliance costs or change the rate of incorrect assessments. Those questions require comparable data that follow the same administrations over time and measure taxpayer outcomes. Future updates to the inventory may make it possible to examine the pace and quality of change. For now, the strongest conclusion is that digital tools have spread across a wide group of economies, while the evidence does not show that transparency, funding and accountability have advanced at the same rate.

Data

Table

Indicators reported in the OECD inventory

Selected coverage and prevalence findings that the report draws from its tax technology inventory covering 108 economies.

Economies covered108
Economies grouped across seven regions104
Administrations requiring digital identity for individuals86%
Administrations requiring digital identity for businesses82%
Administrations developing APIsMore than 80%
Administrations offering public APIsMore than half
Administrations receiving business data beyond pay as you earnAround 60%
Administrations exchanging bulk information across governmentMore than 80%
Administrations reporting AI use49%
AI users reporting limits on AI use85% of AI users
AI users reporting an ethical framework40% of AI users
Administrations reporting a digital transformation strategyAbout three quarters
Administrations reporting strategy period funding58%
Methodology

Values are transcribed from the summary and tables of the OECD report dated 12 November 2026. Underlying responses were collected through the 2024 Global Survey on Digitalisation and were not validated by the OECD or its partners. Percentages describe tax administrations reporting a practice, not individual taxpayer use or measured outcomes. The 85% and 40% AI safeguard figures use administrations reporting AI use as their denominator.

Updated:

Structured coverage

Subjects and places

Subject proposals

Taxonomy version: IPTC Media Topics 2026-Q2

Continue reading

Fiscal Policy

Correction: OECD carbon-price coverage was misstated

The reviewed 2025 report covers 79 countries and reports 44% coverage in 2023 when fuel excise taxes are included, not our claimed 64% across seventy advanced economies.

OECD — Effective Carbon Rates 20252 min read Data included