Tax Administration Digitalisation and Digital Transformation Initiatives

5. Tax rule management and application

Underlying the future vision of tax administration, as described in the 2020 report Tax Administration 3.0: The Digital Transformation of Tax Administration (OECD, 2020[1]), is a shift away from tax rules and calculations being conducted and finalised within the tax administration’s IT environment, to moving the rules and relevant information for tax processing into the taxpayers’ natural systems. This is already seen in many jurisdictions in pay-as-you-earn systems for salaried employees where taxation processes are integrated into the payroll systems used by employers.

With tax systems being complex, tax administrations are increasingly leveraging technology to streamline processes and improve accuracy. For example, more tax administrations are now making use of machine-readable law. This allows legislation to be encoded in a structured digital format that can be processed by the software systems used in taxpayers’ natural systems. This facilitates automated tax calculations, real-time compliance checks, and seamless integration with business and financial systems.

Adopting machine-readable tax rules in appropriate cases can enhance transparency, reduce errors, and improve taxpayer services, while also enabling quicker legislation and policy updates. It can also help to avoid uncertain or inconsistent interpretation of tax legislation. As can be seen in Table 5.1., governments are at different points regarding the use of machine-readable tax law, with around 30% indicating that they publish all tax law in machine-readable format and another 15% indicating that they do this for some tax laws.

Table 5.1. Machine-readable tax law, 2024

Government publishes tax law in machine-readable format
Yes, for all tax laws	Yes, for some tax laws	No
31.5	14.8	53.7

Source: (OECD et al., 2024[2]), Inventory of Tax Technology Initiatives, https://www.oecd.org/content/oecd/en/data/datasets/inventory-of-tax-technology-initiatives.html (accessed on 7 February 2025).

The remainder of this chapter looks at advances in tax rule management, including the use of assurance frameworks to validate the integration of tax rules into third party products and systems, the automation of registration processes, and the role of artificial intelligence.

Assurance frameworks

Alongside improvements in technology and the wide-spread use of software products, an increasing number of third parties is developing software products for tax purposes. To ensure the integrity of the tax system, to build trust and to assist users with selecting the right products, many tax administrations have put in place some form of assurance mechanism. Those mechanisms are meant to ensure that the third-party products apply tax laws in a consistent manner and deliver outputs that can be accepted by the administration.

As can be seen in Table 5.2., around one third of administrations have a dedicated assurance framework, and the vast majority of those that do not have a framework have implemented other mechanisms to ensure that third-party software packages incorporate tax rules correctly:

• 80% of administrations provide technical documentation to enable product development by software providers;

• 77% of administrations implement controls around the authorisation and validation of information sent or received from those software products; and

• 17% of administrations require software providers to follow set requirements in regard to software development cycles.

Only around 6% have no assurance framework or other mechanism.

Table 5.2. Assurance frameworks, 2024

Percent of administrations that have the respective process in place

Assurance framework exists that allows 3rd parties to incorporate tax rules in their software packages and deliver outputs based on those rules that are accepted by the tax administration	Types of approaches taken by the administration where an assurance framework does not exist
Administration provides technical documentation to enable product development by software providers	Administration enforces a software development lifecycle that a software provider must follow to meet set requirements	Administration implements controls around authorisation and validation of information sent / received	Other approaches	No approach
35.2	80.0	17.1	77.1	8.6	5.7

Source: (OECD et al., 2024[2]), Inventory of Tax Technology Initiatives, https://www.oecd.org/content/oecd/en/data/datasets/inventory-of-tax-technology-initiatives.html (accessed on 7 February 2025).

Using trustworthy and reliable and secure software products is key to maintaining trust in the automation of tax rule management. As part of this, some administrations are being increasingly transparent about the software products they approve or recognise.

Table 5.3. shows that a third of administrations publish the names of approved or recognised software products. Around 40% of administrations maintain a register of approved or recognised software products, with slightly more than two thirds of these being public. The table also shows that there are different types of approvals or recognition processes, with the majority of tax administrations using a certification process. Of those that use a certification process, over half have their own certification process, and 18% rely on the certification processes of another government body. Other approaches include an administration endorsement of a private certification process (18%), and a process of self-declaration with the administration then conducting random checks (23%).

Table 5.3. Approved or recognised software products, 2024

Percent of administrations that have the respective process in place

Names of approved or recognised software products published	Register of approved or recognised software products maintained		If yes, type of approval or recognition process
Certification by the administration	Certification by another government body	Administration endorsement of a private certification process	Self-declaration with administration conducting random checks	Other approaches
					Register exists	If yes, register is public
33.3	40.7	68.2	54.5	18.2	18.2	22.7	31.8

Source: (OECD et al., 2024[2]), Inventory of Tax Technology Initiatives, https://www.oecd.org/content/oecd/en/data/datasets/inventory-of-tax-technology-initiatives.html (accessed on 7 February 2025).

Automated registration and deregistration

Taxpayer registration and identification is a key element of an effective tax system as it ensures compliance, enhances trust and transparency, and in some cases can also facilitate access to benefits and grants. It forms the basis for supporting core elements of tax administration work such assessment processes, withholding tax regimes, and third-party reporting and matching.

Automating this process, or parts of it, is an important element of the Tax Administration 3.0 vision (OECD, 2020[1]). Automated registration (and deregistration) at certain events in the taxpayer lifecycle, such as registering a business or registering as a seller on a gig and sharing economy platform, can reduce administrative burdens, increase efficiency and enhance accuracy, while ensuring compliance and building trust.

Table 5.4. illustrates that two thirds of tax administrations can automatically register taxpayers based on data received from third parties, while half can also automatically deregister taxpayers. Some tax systems go further and already allow third parties to automatically register or deregisters taxpayers. However, in most cases this must be with the taxpayer’s consent. Only very few administrations can register or deregister taxpayers without their consent.

Table 5.4. Automated registration and deregistration of taxpayers, 2024

Percent of administrations that have the respective process in place

Based on data received from third parties, administration automatically …		Third party stakeholders can automatically …
Register taxpayers …		Deregister taxpayers …
Registers taxpayers	Deregisters taxpayers	Without taxpayer consent	Only with taxpayer consent	Without taxpayer consent	Only with taxpayer consent
66.7	50.0	3.7	29.6	1.9	20.4

Source: (OECD et al., 2024[2]), Inventory of Tax Technology Initiatives, https://www.oecd.org/content/oecd/en/data/datasets/inventory-of-tax-technology-initiatives.html (accessed on 7 February 2025).

Third party withholding

Withholding regimes support overall tax compliance while significantly reducing burdens for large numbers of taxpayers. This is particularly the case where taxpayers do not need to engage in any post-payment adjustments, for example, where the withholding payment is final or calculated on the correct tax position of the individual taxpayer. This will often require that third parties are provided with information on changes to taxpayer’s individual tax positions so as to be able to withhold the correct amount. As can be seen in Table 5.5. slightly over 20% of tax administrations are now providing such information on an automatic basis.

Table 5.5. Third party withholding of taxes, 2024

Percent of administrations that have the respective process in place

Administration automatically provides information on changes to taxpayer’s individual tax positions to third parties to facilitate withholding of taxes	If yes, …
	Third parties calculate the withholding based on …		Withholding is done as a result of …
A flat tax rate	The individual tax position	Legislative requirements	Voluntary arrangements
22.2	41.7	58.3	91.7	8.3

Source: (OECD et al., 2024[2]), Inventory of Tax Technology Initiatives, https://www.oecd.org/content/oecd/en/data/datasets/inventory-of-tax-technology-initiatives.html (accessed on 7 February 2025).

Artificial intelligence

Artificial intelligence (AI) is being increasingly used by tax administrations to enhance effectiveness and efficiency within the administration, for example on compliance management, and to improve taxpayer services, including through the development of more real-time and personalised processes for interactions with taxpayers. The data from the Inventory of Tax Technology Initiatives shows that 72% of tax administrations use AI (OECD et al., 2024[2]).

There are a range of uses for AI by tax administrations, as can be seen in Table 5.6. The most common is using AI to detect tax evasion and fraud, which is done by three quarters of tax administrations. Other common use cases include as part of risk assessment processes (64%), virtual assistants (59%), aiding officials in making administrative decisions (44%), and making recommendations for actions (41%). Less common uses include ensuring the integrity of tax administration systems/processes (15%), automated provision of personalised information to stakeholders (13%) and dispute resolution (3%). The survey underlying the Inventory of Tax Technology Initiatives (OECD et al., 2024[2]) also asked whether administrations use AI to make final administrative decisions, but none indicated that AI is used in this way.

Table 5.6. Artificial intelligence: Main use cases, 2024

Percent of administrations that use artificial intelligence and have the respective process in place

Automated provision of personalised information to stakeholders	Virtual assistants	Risk assessment processes	Detection of tax evasion and fraud	Assistance of tax officials in making administrative decisions	Making recommendations for actions	Dispute resolution	To ensure the integrity of tax administration systems / processes	Other use cases
12.8	59.0	64.1	74.4	43.6	41.0	2.6	15.4	30.8

Note: Percentages relate to the 72% of administrations that indicated using artificial intelligence.

Source: (OECD et al., 2024[2]), Inventory of Tax Technology Initiatives, https://www.oecd.org/content/oecd/en/data/datasets/inventory-of-tax-technology-initiatives.html (accessed on 7 February 2025).

However, whilst AI undoubtedly has many benefits for tax administrations, its adoption must be carefully managed to mitigate risks around ethical challenges, data privacy and bias. Table 5.7. highlights that a large majority of administrations that use AI (87%) have put limitations in place on its use, and around 40% have an ethical framework regarding the use and application of AI.

Table 5.7. Artificial intelligence: Limitations and ethical framework, 2024

Percent of administrations that use artificial intelligence and have the respective process in place

Limitations exist on the use of artificial intelligence	Ethical framework for the application of artificial intelligence exists
87.2	41.0

Note: Percentages relate to the 72% of administrations that indicated using artificial intelligence.

Source: (OECD et al., 2024[2]), Inventory of Tax Technology Initiatives, https://www.oecd.org/content/oecd/en/data/datasets/inventory-of-tax-technology-initiatives.html (accessed on 7 February 2025).

One of the biggest risks regarding the use of AI is ensuring that it remains objective, with administrations putting various processes in place to ensure this. These processes can either be carried out by the tax administration or by a third party, with Table 5.8. providing a breakdown by approach. The table shows that over 80% of administrations indicated that they themselves carry out reviews of AI input information, the probing and testing of AI responses, and the monitoring of AI outputs. The review of the AI source code is done by half of tax administrations.

Table 5.8. Artificial intelligence: Ensuring objectivity, 2024

Percent of administrations that use artificial intelligence and have the respective process in place

Body that carries out the approaches	Approaches taken by the administration to ensure that decisions by the artificial intelligence are objective
	Reviewing of AI source code	Reviewing of AI input information	Probing and testing of AI responses	Monitoring of AI outputs	Other approaches
Administration	51.3	82.1	84.6	82.1	30.8
Third party	5.1	10.3	10.3	5.1	12.8

Note: Percentages relate to the 72% of administrations that indicated using artificial intelligence.

Source: (OECD et al., 2024[2]), Inventory of Tax Technology Initiatives, https://www.oecd.org/content/oecd/en/data/datasets/inventory-of-tax-technology-initiatives.html (accessed on 7 February 2025).

Innovations

With the management and application of tax rules being automated to a large extent and with vast amounts of data being internally and externally generated, tax administrations continue to look at new and innovative approaches to reduce burdens, provide certainty and ensure compliance.

Table 5.9. highlights some of those innovative approaches. For example:

• 85% of administrations are working with IT systems that automatically identify errors or inconsistencies in the data provided by stakeholders;

• Around 40% of administrations enable the taxpayer or business to test upfront the expected tax treatments of particular scenarios to inform their decision making and allow them to understand their future tax liabilities; and

• Close to 10% of administrations have started using Distributed Ledger Technology in their taxation processes.

Table 5.9. Innovative approaches used as part of tax rule management, 2024

Percent of administrations that have the respective process in place

IT systems are in place that automatically (i.e. without human intervention) identify errors or inconsistencies in the data provided by stakeholders	IT system enable the taxpayer or business to test upfront tax treatments of scenarios to inform their decision making and liabilities	Distributed Ledger Technology (DLT)
DLT used in administration’s taxation processes	If yes, administration uses smart contracts embedded within DLT
85.2	42.6	9.3	60.0

Source: (OECD et al., 2024[2]), Inventory of Tax Technology Initiatives, https://www.oecd.org/content/oecd/en/data/datasets/inventory-of-tax-technology-initiatives.html (accessed on 7 February 2025).

References

[1] OECD (2020), Tax Administration 3.0: The Digital Transformation of Tax Administration, OECD Publishing, Paris, https://doi.org/10.1787/ca274cc5-en.

[2] OECD et al. (2024), Inventory of Tax Technology Initiatives, https://www.oecd.org/content/oecd/en/data/datasets/inventory-of-tax-technology-initiatives.html (accessed on 7 February 2025).