Artificial intelligence is unlikely to eliminate the accounting profession, but it can significantly reduce the cost of delivering basic accounting services. In Poland, that shift is being accelerated by KSeF, the national e-invoicing system. Accounting firms that continue to rely heavily on manual document processing may face pressure on prices, staffing and margins, while highly automated competitors could serve far more clients with similar resources.
KSeF turns invoices into structured data
KSeF is particularly important in Poland because it changes the basic nature of invoice processing. Large taxpayers have been required to issue invoices through the system since 1 February 2026, with the obligation extended from 1 April 2026 to most other businesses, subject to transitional rules for the smallest firms. Receiving invoices through KSeF has been mandatory since 1 February.
Instead of starting with a PDF or paper document that has to be manually read and entered into software, accounting systems can increasingly receive structured data directly. The workflow can therefore move from invoice → email → PDF → employee → accounting software toward KSeF → integration → accounting system → rules and AI → exception review → posting.
Accountants may review exceptions instead of every transaction
If a company receives the same types of leasing, rent, telecoms, hosting or other recurring invoices each month, a system can compare new documents with historical records, verify counterparties, amounts, VAT rates, bank accounts and previous accounting treatment.
In such a model, the accountant does not need to manually search for problems across hundreds or thousands of normal transactions. The system should flag unusual cases that genuinely require professional judgement.
Private AI environments may matter more than public chatbots
Financial data makes security a central issue. For many companies, the target model is unlikely to involve copying client data into a public chatbot. A more practical approach is a private or controlled AI environment with access only to defined systems, documents and sources.
Such an assistant could use a client’s accounting history, chart of accounts, procedures and previous decisions to detect anomalies, compare periods, prepare reports and identify documents requiring review.
A Wolters Kluwer survey conducted in late 2025 among 581 accounting, payroll and tax professionals found that more than 80% were already using AI tools. Respondents most often pointed to automation, time savings and better analysis as benefits, while data security and system integration remained important challenges.
Lower costs will put pressure on prices and margins
If automation materially reduces the cost of serving a client, the first movers may initially keep a larger share of the savings as higher margins. Over time, however, competition is likely to pass part of those savings on to clients.
That suggests growing price pressure in basic, repetitive transaction processing. But it does not necessarily mean a race to the bottom. Accounting firms can also use automation to provide more value for the same fee through cash-flow analysis, anomaly alerts, receivables monitoring, tax forecasts or management reporting.
Who can benefit most?
Software providers are among the clearest potential winners because the largest value may lie not in the language model itself, but in secure integration with KSeF, banks, ERP systems, payroll, payments, tax databases and company-specific historical data.
Large automated accounting firms could also gain economies of scale if one employee can supervise a larger client portfolio. Smaller technology-focused firms may benefit as well, while businesses should ultimately see either lower fees or a broader service offering.
Will the number of accounting firms fall?
A gradual consolidation of the market appears more plausible than the sudden disappearance of accounting firms. The most exposed business models are those based mainly on receiving documents, entering them, posting transactions and informing clients about tax amounts.
Tasks involving unusual situations, risk assessment, interpretation, responsibility and communication with business owners are much harder to automate.
AI may therefore eliminate a large share of the work clients currently pay accountants to perform manually, rather than eliminate accountants themselves.
What accounting firms should do now
The first step should not be buying another chatbot. Firms should map their entire client-service process and divide each activity into three groups: fully automatable, suitable for AI or system support under human supervision, and requiring specialist judgement.
A likely target model is: KSeF + bank data + accounting software + automation + a secure AI environment + an accountant acting as an exception controller and adviser.
Sources: Polish Ministry of Finance information on the mandatory KSeF timetable; Wolters Kluwer research on AI in accounting and payroll.
