The digitalisation of tax administration: Why data is becoming more important than forms
What you can expect from this article
Time and again, I hear about experiences like this or similar ones: a piece of information is already on file in the accounts department. Nevertheless, for tax reporting purposes, it is transferred once again into a spreadsheet, cross-checked against other data and finally entered into a form. Admittedly, this is digital in a way. But is it efficient? A new OECD study shows how tax administrations are changing internationally. The focus is shifting away from electronic forms alone and increasingly towards the question of how data flows between systems. For businesses and tax firms, this is a good opportunity to take a closer look at their own processes.
‘Digitally submitted’ does not yet mean ‘digitally conceived’
The electronic submission of tax returns and other documents was an important step. After all, it is certainly convenient when documents no longer need to be printed out, signed and sent by post. But what actually happens before we click „Send“? Information is often gathered from various applications, transferred into spreadsheets and checked several times. Sometimes the same data is even re-entered, even though it has long been available elsewhere. This takes time, creates opportunities for error and is still a long way from fully digitalised processes.
In its study published on 12 November 2026, the OECD describes a development that addresses precisely this issue. Tax administrations should integrate more closely with existing business processes so that information can be processed, as far as possible, at the point where it is generated. In doing so, the organisation is building on its vision of „Tax Administration 3.0“. The analysis draws on data from 108 economies. More than 80 per cent of the tax administrations surveyed are developing application programming interfaces, or APIs. Over half have already made such interfaces publicly available. And around 60 per cent receive data directly from taxpayers’ business systems, going beyond payroll tax processing. This points the way forward, even though the development is at different stages of advancement across regions.
An interface alone does not solve a data problem
Now, one might say: ‘Brilliant, we’ll just connect all the systems together and the rest will run automatically.’ As nice as that would be, it’s unfortunately not that simple. Whilst an interface can transfer data, it cannot ensure that this data is complete, accurate and correctly assigned. This becomes clear, for example, with a tax return, which requires information from the accounts, a list of shareholdings and internal documentation. If a company is referred to differently in each system, or if the time periods do not match up, even the fastest connection is of little help. In that case, the error simply travels automatically from one system to the next. Whilst this is consistently digitised, it is probably not the result we are looking for.
For me, therefore, automation does not begin with data transfer, but much earlier: with the question of what information is required, where it comes from and who is responsible for it. Unambiguous data structures, plausibility checks and traceable corrections are not tedious additional tasks. Instead, they lay the foundations for digital processes to function reliably. The OECD summarises an important concept under the principle „Collect once, use many times“: capture information once wherever possible and use it multiple times. The report focuses in particular on data exchange within the public administration and between public authorities. However, the concept can also be applied to collaboration within companies and law firms. If information is available in a structured format and has been professionally verified, it does not need to be gathered from scratch again for every purpose.
Data usage and data security go hand in hand
The more easily information flows between applications, the more important the question of who is authorised to access it becomes. After all, not everyone who needs a piece of information should be able to modify or transmit it. And in the event of an enquiry, it should be possible to trace who made a change or granted authorisation. The OECD study shows that this issue is also relevant for tax authorities: 95 per cent of the authorities surveyed state that they monitor user access to data. In my view, this means that for businesses and law firms, authorisations, responsibilities and documentation must be incorporated into process design from the very outset – not only once all systems are already interconnected.
Artificial intelligence also plays a role in the study. According to the OECD, almost half of the tax administrations surveyed use AI to support compliance with tax regulations and improve services. Common areas of application include fraud detection and risk assessment. However, this does not mean that tax decisions are already being made autonomously everywhere. AI can also be helpful for businesses and law firms, for example to identify anomalies in data more quickly or to analyse documents in a more targeted manner. For me, it is crucial that results remain verifiable and that responsibility is clearly defined. AI can support people. But it should not lead to a situation where, in the end, nobody can explain how a result was arrived at.
What businesses and law firms can do now
The OECD study does not establish any new German reporting requirements. Instead, it highlights international developments and thereby provides impetus for Germany’s own work. It is important to note that the underlying data is derived from self-reported information provided by the economies surveyed and has not been independently validated by the OECD. It is therefore not possible to draw direct conclusions about the state of German tax administration from the international figures. Nevertheless, it is worth taking a look at our own processes. Where is data exported from one system, processed in Excel and then re-entered elsewhere? What information is required for multiple tasks? And who checks, updates and approves it? Answering these questions often provides the first starting points for reducing data disconnects and creating more traceable processes.
At hsp, we have been exploring for many years how such gaps in data continuity can be reduced. This is because tax-related information is rarely generated in just one place and is rarely processed by just one person. Companies, law firms and other stakeholders must work together, coordinate information and document the results. With Opti.Tax and the relevant modules, we support, amongst other things, structured documentation and financial reporting. The optional Opti.Tax Cloud takes collaboration between law firms and clients to a new level, for example when exchanging documents and managing tasks. Which specific workflows and interfaces can be utilised depends on the particular use case.
hsp News April 2023: CbCR in Opti.TaxDuring the live broadcast on 28 April, hsp CEO Paul Liese once again spoke about the latest projects and developments at hsp. He began by expressing his gratitude for the many constructive, entertaining and exciting encounters at the StB EXPO in Munich. Paul then briefly pointed out that tax consultancy firms will not be able to avoid the issue of money laundering.
Low-code/no-code: Advise, automate, reap the benefitsIn software development, certain processes are constantly repeated. In response to this need, development platforms have emerged where developers no longer need to write many, if any, lines of code: low-code and no-code. Instead of using a text-based programming language, they work with visual tools.








