Tax audits in Poland: how KSeF and JPK_CIT are changing tax supervision

Tax audits in Poland are entering a new stage of digitalisation. For companies, this means a change that goes much deeper than the implementation of the National e-Invoicing System (KSeF) in Poland or the preparation of systems for Standard Audit File for Corporate Income Tax reporting (JPK_CIT). The Polish tax administration is increasingly able to analyse taxpayers’ data before making formal contact with the company, and the decision to initiate verification activities may result from earlier comparisons of information from multiple sources.

In practice, a tax audit is no longer only an event that starts when a formal notice is delivered. More and more often, it is preceded by data analysis, identification of unusual patterns, assessment of links between entities and preliminary identification of risk areas. Only at a later stage may the tax authority contact the taxpayer with specific questions, a request for explanations or a demand to provide documents.

For companies, the key practical conclusion is clear: tax correctness will not be assessed only on the basis of the final tax return. The consistency of data across the entire process will become increasingly important — from the invoice, through accounting entries, to reporting in SAF-T structures, KSeF, JPK_CIT and the financial statements.

---

Digitalisation of taxes is changing the way companies should think about tax audits

For many years, the main point of reference in Polish tax audits was documentation: invoices, contracts, registers, accounting printouts and explanations prepared by the taxpayer. Today, this model is gradually being replaced by data-based supervision.

This does not mean that documents are losing their importance. They are still necessary to demonstrate the course of a transaction and justify tax settlements. What is changing, however, is the order of actions.

The tax authority may increasingly analyse data first, and only later ask for documents confirming or explaining the picture that emerges from that data.

This approach is possible thanks to the development of electronic reporting in Poland. Increasingly important sources include, among others:

• Standard Audit File for VAT (JPK_VAT),
• SAF-T files submitted on request,
• electronic financial statements,
• the ICT System of the Clearing House (STIR),
• the VAT taxpayers white list,
• the National e-Invoicing System (KSeF),
• SAF-T structures for income taxes, commonly referred to in practice as JPK_CIT.

In this model, the tax authority can compare data from different sources and check whether they form a logical whole. If a transaction appears on an invoice, it should be possible to link it with the accounting entry, VAT register, income tax calculation, payment and business documentation.

---

KSeF and JPK_CIT are not just another technical obligation

Many companies treat the National e-Invoicing System (KSeF) in Poland and JPK_CIT primarily as implementation projects. This is natural, because they require changes in systems, document workflows, accounting processes and reporting. From the perspective of tax audits in Poland, however, their significance is much broader.

KSeF increases the availability of invoice data

A structured invoice contains information in a format that can be automatically read, compared and analysed. Data concerning the parties to the transaction, amounts, VAT rates, dates, document numbers and other invoice elements may become the starting point for further verification.

JPK_CIT increases the importance of accounting data in corporate income tax

In practice, JPK_CIT involves reporting accounting books and records in SAF-T structures for income taxes, including, among others, JPK_KR_PD for taxpayers maintaining accounting books. Accounting and tax data will have to be presented in a more structured way, and its internal logic will become easier to verify.

The biggest risk is therefore not only whether the file is generated in the correct technical format. What matters much more is whether the data included in the file is consistent with other information available to the tax authority.

For example, a company may generate a file correctly from a technical perspective but still face problems if:

• accounting data does not match the adopted mapping rules,
• the method used to classify costs cannot be reconstructed,
• discrepancies between systems do not have a clear justification,
• adjustments are made outside the main system without proper documentation,
• similar transactions are recorded differently without a clear reason.

In such situations, technical system readiness is not enough. Companies also need accounting and tax processes aligned with Polish regulations and local reporting practice. Local accounting services in Poland and tax advisory in Poland can help verify whether reporting data is consistent, complete and defendable in the event of a tax audit.

---

A new model of supervision: the tax authority sees more, faster and earlier

The digital model of tax supervision in Poland is based on the assumption that data can reveal risks faster than a traditional review of documents. The Polish tax administration may analyse data submitted by taxpayers, information from public registers, financial data, banking data, tax settlement history and results of previous verification activities.

In practice, this means that the tax authority may identify certain risk signals even before contacting the company, for example:

• unusual changes in the structure of sales or costs,
• repeated corrections relating to similar transactions,
• differences between invoice data and accounting records,
• transactions with higher-risk counterparties,
• unusual financial flows,
• discrepancies between VAT, CIT and accounting data,
• lack of continuity in the classification of similar events,
• personal or capital links between participants in a transaction.

This does not automatically mean that every such situation is a tax error. In business, many events have an economic justification, and differences between VAT, CIT and accounting often result from different legal rules.

The problem arises when the company is unable to explain them quickly, consistently and on the basis of data.

---

Remote tax audits are changing communication with the tax authority

An increasing part of the tax authority’s activities may be conducted without the physical presence of inspectors at the taxpayer’s premises. Data can be submitted electronically, and the authority can analyse it remotely. From the company’s perspective, this changes the dynamics of a tax audit.

In the traditional model, an audit often began with a broad request for documents. In the digital model, questions may be much more precise.

The tax authority may ask, for example, about:

• a specific group of invoices,
• selected transactions with a given contractor,
• a discrepancy between two data sources,
• the method used to classify a specific type of cost,
• the reason for an adjustment,
• the moment when revenue or cost was recognised,
• the consistency of KSeF data with accounting records.

A company that does not have well-organised data may find it difficult to respond even if the tax settlement itself was generally correct. The risk does not then result from the transaction itself, but from the inability to reconstruct and explain it quickly.

That is why preparation for a tax audit in the digital era should cover not only tax documentation, but also:

• data quality,
• invoice workflow,
• the accounting process,
• descriptions of system rules,
• information archiving,
• the history of changes and approvals.

For companies using multiple systems or operating internationally, local expertise is essential. getsix® helps businesses align reporting processes, document workflows and communication with Polish tax authorities through accounting outsourcing in Poland and ongoing tax advisory in Poland.

---

Algorithms do not replace tax audits, but they may indicate risk areas

In discussions about tax digitalisation, one question often appears: will algorithms decide whether a company is subject to a tax audit? From the entrepreneur’s perspective, another issue is more important: analytical tools may indicate to the tax authorities areas that require closer verification.

Algorithmic data analysis does not determine that the taxpayer has breached the rules. However, it may reveal an unusual pattern, discrepancy or connection that will then be assessed by tax administration officials. In this way, data becomes a filter that helps the authorities select cases for further action.

For businesses, this means that elements previously treated as technical details are becoming more important, including:

• correctness of contractor data,
• consistency of descriptions,
• consistency in the use of accounting accounts,
• up-to-date mapping,
• consistency of dates,
• history of corrections,
• completeness of the audit trail.

A formal or system error may not cause a significant tax arrear, but it may increase the likelihood of questions from the tax authority. In the digital era, data quality becomes part of tax risk management.

---

Main sources of risk for companies

The new audit model does not create completely new tax obligations detached from existing Polish tax regulations. However, it reveals problems that have already existed in many organisations but were previously less visible.

Discrepancies between systems

In many companies, tax data is created in several places. Some information is stored in the ERP system, some in sales systems, some in document workflow tools, some in spreadsheets and some in supporting systems.

If data from these sources is not reconciled at transaction level, differences may arise in relation to:

• dates,
• amounts,
• document numbers,
• contractors,
• VAT rates,
• accounting accounts,
• tax categories.

In the traditional model, such differences were often corrected only at the end of the period. In the digital model, they may be noticed much earlier.

Manual corrections and spreadsheets outside the system

A second source of risk is processes based on manual reconciliations. Spreadsheets are not a problem in themselves, but they become risky when they contain key tax logic that cannot be easily verified.

If the calculation, correction or tax classification depends on a local file and the knowledge of one person, the company may have difficulty demonstrating why the data was processed in a particular way.
In a tax audit, not only the final number matters, but also the ability to reconstruct the process that led to it.

Errors automatically repeated by systems

Automation reduces the risk of manual errors, but it increases the importance of correct configuration. One error in account mapping, a tax rule or code assignment may be repeated across hundreds or thousands of transactions.

Such an error may become systemic. When analysing data, the tax authority may notice a recurring pattern and ask about its cause.

That is why every automation in accounting and tax should be:

• tested,
• documented,
• approved,
• periodically reviewed.

Inability to reconstruct decisions

In a tax audit, one question is often crucial: why did the company record a given transaction in this particular way?

If the answer is based only on an employee’s memory or informal practice, the taxpayer’s position is weaker.

A company should have procedures, process descriptions, change history and evidence of approvals. This applies especially to areas that may be subject to disputes, such as:

• tax-deductible costs,
• VAT settlements,
• corrections,
• transactions with related parties,
• international settlements,
• classification of unusual benefits or services.

---

Data must be consistent not only at the end of the period

Many organisations are used to tax data being organised at the end of the month, quarter or year. In the digital supervision model, this approach may be insufficient.

The tax administration may analyse data continuously and compare how it changes over time. Therefore, a company should know not only what the final tax settlement result was, but also how that result was created.

Important elements include:

• data versions,
• correction history,
• change logs,
• the ability to reconstruct the status as at a specific date,
• justification of differences between source data and the report.

This is particularly important in larger organisations, where data passes through many stages: from the sales system, through document workflow, approval, accounting, tax settlement, management reporting and financial reporting.

If manual changes are introduced at every stage but are not documented, the company may find it difficult to defend the consistency of its data.

---

How should companies prepare for digital tax audits in Poland?

Preparation for the new audit model should begin with a change of perspective. The point is not only to submit the correct SAF-T file or handle an invoice in KSeF. The point is to ensure that tax data is organised, consistent and justifiable.

1. Review data sources

The first step should be to determine where the data used for tax settlements comes from. The company should identify all systems and tools that affect VAT reporting, CIT reporting, accounting books and financial reporting.

It is worth checking:

• which data is source data,
• where data is modified,
• who has access to it,
• which parts of the process are performed manually,
• which data is used for tax reporting.

2. Reconcile data between systems

The next step is to compare data between systems. It is not enough to check whether final totals match. It is much more important to reconcile data at the level of individual transactions.

The company should know whether an invoice visible in KSeF has been:

• correctly recorded in the accounting books,
• linked to the correct contractor,
• matched with the VAT register,
• assigned to the appropriate tax category,
• correctly included in the CIT settlement.

3. Review accounting and mapping rules

A review of data mapping is particularly important. This includes the assignment of accounting accounts, tax categories, VAT rates, markings and reporting fields.

If mapping is outdated, inconsistent or undocumented, it may lead to errors repeated in subsequent periods. In the case of JPK_CIT, this will be one of the key risk areas.

4. Limit manual exceptions

The company should check which parts of the process are performed outside the system and why. Not every activity can be fully automated, but exceptions should be described, approved and controlled.

The more key tax decisions are made in spreadsheets and informal arrangements, the more difficult it will be to demonstrate the repeatability and reliability of the process.

5. Build an audit trail

One of the most important elements of preparation for a tax audit is the ability to trace data from the source document to the tax report.

Such an audit trail should include:

• the source document,
• contractor data,
• document approval,
• accounting entry,
• tax classification,
• possible corrections,
• reporting,
• archiving.

The company should be able to show how a given transaction passed through the process and why it was recorded in a particular way.

6. Conduct test reporting

In the case of JPK_CIT, it is worth conducting test reporting on historical data. This makes it possible to check whether the process works in practice, not only whether the system is technically ready.

A test may reveal: missing data, inconsistent rules, incorrect mapping, problems with fixed asset records, difficulties in reconstructing the history of changes.

7. Simulate questions from the tax authority

The final step may be to look at the data from the tax authority’s perspective. It is worth asking questions that may arise during a tax audit:

• Why were these transactions classified in this way?
• What causes the differences between VAT and CIT?
• Why is data in different systems not identical?
• Who approved the correction?
• Are similar transactions recorded consistently?
• Can the company reconstruct data from the period covered by the audit?

Such an analysis allows the company to prepare explanations before the tax authority asks for them.

---

The role of management and the finance department

Digital tax audits are not only an accounting department issue. They also concern the management board, CFO, IT department, sales department, purchasing department and people responsible for operational processes.

Tax data is created much earlier than at the moment of accounting. Its quality depends on how contracts are concluded, contractor data is entered, services are described, invoices are approved, costs are assigned, systems are managed and access rights are controlled.

Management should be aware that, in the new model, tax risk may result not only from an incorrect interpretation of regulations, but also from a disorganised process.

That is why preparation for the National e-Invoicing System (KSeF) in Poland, JPK_CIT and tax audits in Poland should be treated as an organisational project, not only a technical task.

---

Why is it worth acting before a tax audit?

A company’s greatest advantage is the ability to identify problems before the tax authority does. If an organisation detects an inconsistency itself, it can assess its cause, correct the data, supplement documentation or change the process for the future.

Acting only after receiving a request from the tax authority usually means greater time pressure and less control over the situation. The company must then not only explain the data, but often also reconstruct a process that was not properly documented earlier.

For this reason, preparation for digital tax audits should include a periodic review of data and processes. This is not about a one-off audit, but about a permanent mechanism for controlling the quality of tax data.

---

Tax audits in Poland in the era of the National e-Invoicing System (KSeF and JPK_CIT will be increasingly based on data analysis. The Polish tax administration is gaining the ability to compare information faster, identify inconsistencies and select risk areas even before formal contact with the taxpayer.

For entrepreneurs, this means the need to change their approach to tax settlements. It is no longer enough to prepare a correct tax return or generate a file in the required format. Data must be consistent, traceable and justifiable at every stage of the process.

The best-prepared organisations will be those that treat KSeF and JPK_CIT as part of a broader tax data management system. Key elements will include: quality of source data, process control, reduction of manual exceptions, documentation of decisions, the ability to explain discrepancies quickly.

In the digital supervision model, companies that do not wait for a tax audit but analyse their own data in advance from the perspective of potential questions from the tax authority will gain a clear advantage.

---

If you have any further questions or require additional information, please contact your business relationship person or use the enquiry form on the HLB Poland website.

Contact us »

***