TP auditing in the tech environment

International groups and tax administrations are facing an incredible acceleration of the pace of digitalisation and the evolution and adoption of new technologies. This is an opportunity for taxpayers to introduce new tools and perform tax functions in a new way. This is also a challenge, as this IT evolution is leading to an increase in accessible information, a change in the tools available to tax auditors, and, of course, from time to time, certain issues with accessing information and sharing useful data in a mutually acceptable format (when information is requested, during a compliance process, and tax audits).

The impact on taxpayers’ relationships with tax administrations and on tax controversy protocols, approaches, and actions is, and will be, increasingly significant, particularly during tax audits. AI is a key concern for groups seeking to understand how tax authorities may use artificial intelligence and how the legal framework regarding AI and taxpayer information access will evolve. In this rapidly changing environment, groups can still adapt their strategy, maintain strong relationships with tax administrations, and prepare for possible procedural changes – both to explain past activities and to obtain greater tax certainty for the future.

The following developments highlight the main concerns at present and help explain how tax administrations are responding to these IT evolutions and taking better advantage of these new technology opportunities.

Tax administrations are continuing to adjust to the AI and IT evolution, to stay close to taxpayers’ technical evolutions and retain access to useful information, as well as to capture efficiency improvements in their organisation and help their auditors to be more accurate during tax audits.

The example of e-invoicing (which provides a new way for tax administrations to manage VAT collection and to gather information that will likely be useful for transfer pricing (TP) purposes) is probably the most visible element of this evolution. Such a radical change is obviously more complicated to introduce in a tax audit procedures environment.

Tax auditors now have access to much more information:

• From the tax administrations’ internal documents and figures, coming from taxpayers’ returns and provided by treaty partners under the exchange of information procedures;

• From public databases and open-source figures (taxpayers’ public reports, blogs, and posts published by official groups’ marketing departments, as well as by employees); and

• From groups and audited companies, as they are collecting more accurate internal information.

In this new environment, tax auditors can have a better understanding of what the group is doing, and they must use new tools to make all this information as useful as possible.

At the same time, tax administrations in most countries must adapt to public budget constraints:

• Resources are becoming scarcer, with fewer staff and reduced financial capacity; and

• Governments are seeking to improve budgetary efficiency by increasing revenues and reducing tax gaps and financial leakage.

Experience shows that in this rapidly evolving environment, tax audit departments are making greater use of available information and IT tools. They are seeking to adapt as effectively and quickly as possible.

However, it may be complicated for tax administrations to follow the rhythm of this quick evolution and the pace of IT changes, as they may lack IT specialists.

Moreover, AI and IT procedures must be capable of being integrated into audit procedures and recognised within the applicable national legal framework. Because this framework is often relatively rigid, change may occur more slowly than governments expect, or tax administrations require. As a result, administrations may need to adopt transitional solutions.

For example, some administrations have sought to introduce fully paperless procedures for exchanging official documents. However, pending more developed and comprehensive IT solutions, they often had to rely on basic tools already available on the market, as a first step.

Sometimes, slow implementation can be explained by well-known questions connected to the strength of the audit procedure, privacy of information, and confidentiality:

• Tax audit procedures are designed to ensure regular exchanges between tax audit departments and taxpayers, while preserving a record of communications and applicable deadlines. Evidencing dematerialised procedural steps and securing proof can be difficult and may slow the deployment of IT solutions.

• Automated access to large volumes of information may also raise concerns because it may not distinguish between information genuinely relevant to the audit and information that should remain outside its scope (legally prohibited, unrelated to the TP investigation, or private).

• As in other organisations, the use of certain IT solutions may raise questions about access to information processed by AI. This can create concerns about information leaks and, for some governments, data sovereignty.

Digitalisation is gradually transforming tax audits from document-based verification exercises into reviews focused on data, systems, and processes. Tax authorities increasingly examine the full life cycle of tax-relevant data: where data is generated, how it is processed, which systems and interfaces transform it, and how it ultimately flows into tax returns, financial statements, and TP documentation. The focus therefore shifts from individual records to data consistency, interface reliability, and tax data governance.

In the traditional audit model, tax authorities primarily reviewed individual documents, accounting entries, and explanations provided by the taxpayer. In a digital audit environment, relevant tax facts are increasingly assessed by reviewing ERP systems, sub-ledgers, document management systems, email archives, and data interfaces. Accordingly, the focus is shifting to earlier in the data life cycle: at the point where data is created, modified, transferred, aggregated, and interpreted.

For TP, this development is particularly relevant. TP positions are no longer tested only against documentation reports, intercompany agreements, and benchmark studies. They can increasingly be cross-checked against internal and external data sources, including accounting data, (public) country-by-country reporting information, DAC6 disclosures, tax-relevant communication, publicly available information, and information exchanged between tax authorities. This creates opportunities for more targeted audits but raises concerns regarding proportionality, legal relevance, data quality, confidentiality, and the risk that statistical patterns may be mistaken for evidence.

AI and advanced analytics can support risk-based tax audit case selection. In TP cases, potential risk indicators may include persistent losses, lower margins than other taxpayers in the same industry, inconsistencies within submitted relevant information (transaction matrix or documentation), or other available information. For this purpose, tax authorities may request structured accounting data exports, general ledger data, ERP reports, interface documentation, contracts, invoices, workflow logs, and, where tax-relevant, electronic communication (emails). Digital access may include direct access or the provision of relevant ERP data (SAP-DART files).

A first step is assessing whether recorded data is complete and internally consistent. This may involve interface validation between source systems, general and sub-ledgers, and the review of mapping logic for common error sources such as incorrect signs, wrong tax rates, or duplicate postings. As a second step, data may be statistically analysed to identify structures, trends, outliers, and unusual patterns. Process mining can reconstruct actual process flows from event data and show whether the process as performed corresponds to the documented process.

For TP tax audits, this means that the documented functional and risk profile can be tested against actual conduct. Emails, approval workflows, pricing decisions, customer contacts, and escalation paths may be reviewed to assess which entity performs key functions or exercises control over risk. This can make audits more evidence-based but also increases the risk of premature conclusions if data is interpreted without sufficient business context.

Digitalisation changes the nature of errors. In an analogue environment, errors are often isolated and transaction specific. In a digital environment, digital errors are frequently systemic. An incorrect interface mapping or parameter, flawed allocation key, or automated processing rule may affect thousands of transactions across several periods. In TP, digital audits may uncover systemic errors in cost bases, allocation keys, segmented P&Ls, and royalty fee calculations.

A practical TP AI use case currently developed in Germany is network analysis of tax-relevant mass business communication, including emails and potentially other electronic correspondence. Header data can reveal communication networks, key actors, hubs, and unusual contact patterns. This may be used to test whether the contractual allocation of functions and risks aligns with actual decision-making. For example, if a foreign principal is documented as the entrepreneur, but German employees appear as central communication hubs for pricing, product strategy, or customer relationship management, the documented functional and risk profile may be challenged. However, communication density alone does not prove entrepreneurial control. Any reassessment must be based on qualitative review and corroborating evidence.

There are several limitations and challenges to digital tax audits:

• The scope of data requests remains limited by tax relevance, proportionality, and statutory retention obligations. Technical accessibility may not automatically create an obligation to provide all available data.

• A lack of objectivity – data and algorithms are not neutral. Algorithmic risk selection should be understood as a screening mechanism, not as proof. Correlations, clusters, and anomalies do not replace factual investigation or legal analysis.

From a legal perspective, human oversight, source validation, and reproducibility of results remain indispensable. This is because the fiscal authority generally bears the burden of proof for tax adjustments and must be able to substantiate its findings not only during the tax audit but also in subsequent litigation or mutual agreement procedures.

Tax teams must adapt to the risks and opportunities presented by an expanding suite of digital tools available to both their team and tax authorities, and transition from a reactive ‘document and defend’ approach to a more proactive model of tax governance, data readiness, and audit preparedness. Digitalisation not only increases the amount of information available but it also changes how that information is reviewed, connected, and interpreted. This means tax positions must be supported not only by documentation but also by consistent data, reliable systems, and evidence of actual conduct.

This urgent adaptation is accelerating the trend of tax leadership, moving beyond compliance and reporting, and towards a business advisory role working alongside the group C-suite to influence policies, business strategies, and investment decisions.

A key priority is to establish a single, coherent version of the group’s TP facts. TP documentation reports, intercompany agreements, country-by-country reporting, management reporting, ERP data, and operational workflows should tell the same story. Inconsistencies between the documented TP model and the actual data environment are likely to become more visible as tax authorities use analytics to compare information across filings, jurisdictions, and data sources.

This is particularly important after business reorganisations, M&A, ERP migrations, or changes to supply chains. In such situations, different systems, reporting structures, and local practices may create competing versions of the facts. Tax teams should work closely with finance, IT, legal, and business stakeholders to ensure TP policies are not only documented but properly implemented and reflected in transactional data.

Tax data governance will become a core element of digital audit readiness. Groups should understand where relevant tax data originates, how it is transformed, which systems process it, and how it can be reconciled to financial statements, tax returns, and TP documentation. This includes clear ownership of data, documented extraction logic, data dictionaries, interface descriptions, and controls over allocation keys, cost bases, royalty bases, and segmented P&L statements.

The objective is not to create perfect systems but explainable, reproducible, and defensible ones. If tax authorities identify anomalies, the taxpayer should be able to explain whether it reflects genuine business facts, system artefacts, mapping issues, or errors. Without such transparency, even technically harmless inconsistencies may lead to lengthy audit discussions.

TP documentation will increasingly be tested against operational evidence. A functional and risk analysis should be consistent with approval workflows, pricing decisions, customer contacts, escalation procedures, and management reporting depicted in electronic communication.

This does not mean that every communication or workflow is decisive. Data patterns, communication networks, and process traces are only indicators. Taxpayers should be prepared to explain how decision-making actually works and how the documented TP model is supported by reliable evidence.

Technology is essential to manage the increasing volume and complexity of tax-relevant data. Taxpayers should consider tools supporting documentation, data reconciliation, audit readiness, and anomaly detection. However, technology alone is not sufficient. Taxpayers need people with the skills to understand data flows, challenge analytics outputs, and translate technical findings into tax-relevant explanations.

To keep pace with this fast-moving technical landscape, taxpayers need to strengthen IT capabilities and their TP operating model. Dedicated technology solutions, including AI-based tools developed by Deloitte, can support different stages, such as data management, documentation, compliance, audit readiness, and dispute support, including litigation and tax raids (subject to applicable professional independence rules for restricted clients).

Greater transparency can support credibility, but it must be managed carefully. Taxpayers should assess information requests against tax relevance, proportionality, confidentiality, and retention obligations. Technical accessibility does not mean that all available data must be provided. However, taxpayers should avoid appearing defensive where the requested information is relevant and can be explained.

A balanced disclosure strategy requires preparation before the audit begins. Audit-ready files, reconciliations, and supporting evidence can help respond efficiently while maintaining control over the narrative.

Human relationships will remain critical and may become even more important as data transparency and analytical capabilities increase. Digital tools allow tax authorities to review larger volumes of data and identify potential anomalies more efficiently, but trust, credibility, and constructive dialogues remain key success factors.

In practice, an auditor’s confidence in the taxpayer’s data, processes, and governance framework will influence the depth and focus of audits. Transparency should therefore not be viewed solely as a technical challenge but as an opportunity to strengthen credibility and foster cooperative relationships with tax authorities.

This is where the human element remains essential. Beyond technology and tools, experienced advisers help taxpayers interpret data, prepare explanations, manage procedural risks, and maintain a constructive dialogue with tax authorities. Deloitte’s expertise can be critical in helping groups combine data-driven audit readiness with sound judgement, practical experience, and effective controversy management.

Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited (DTTL), its global network of member firms, and their related entities (collectively, the “Deloitte organization”). DTTL (also referred to as “Deloitte Global”) and each of its member firms and related entities are legally separate and independent entities, which cannot obligate or bind each other in respect of third parties. DTTL and each DTTL member firm and related entity is liable only for its own acts and omissions, and not those of each other. DTTL does not provide services to clients. Please see http://www.deloitte.com/about to learn more.

This communication contains general information only, and none of Deloitte Touche Tohmatsu Limited (DTTL), its global network of member firms or their related entities (collectively, the “Deloitte organization”) is, by means of this communication, rendering professional advice or services. Before making any decision or taking any action that may affect your finances or your business, you should consult a qualified professional adviser.

No representations, warranties or undertakings (express or implied) are given as to the accuracy or completeness of the information in this communication, and none of DTTL, its member firms, related entities, employees or agents shall be liable or responsible for any loss or damage whatsoever arising directly or indirectly in connection with any person relying on this communication. DTTL and each of its member firms, and their related entities, are legally separate and independent entities.

© 2026. For information, contact Deloitte Global.