How is AI changing the skills expected of professionals in your jurisdiction?
Artificial intelligence is becoming embedded within German tax advisory practices and, while it is unlikely to transform every office overnight, its impact on the profession is already significant.
The direction of travel shows that routine tasks will increasingly be handled by technology, while judgement and accountability remain firmly with human professionals.
AI has the potential to automate repetitive activities such as processing accounting data and reviewing documents, creating capacity for the services clients value most: strategic advice, forward-looking tax planning and trusted support in key business decisions.
Successfully integrating AI, however, requires far more than selecting the right software. Professionals must develop a broader skill set that extends beyond technical tax knowledge. Team members need to be capable of critically assessing AI-generated outputs, identifying errors, understanding the limitations of automated analysis and recognising when human judgement must take precedence. Digital literacy, structured data management, a working understanding of the GDPR and the EU AI Act and a healthy scepticism towards seemingly authoritative AI-generated results are becoming essential professional competencies. Put simply, the stronger the professional foundation, the more effectively and safely AI can be deployed.
At the same time, AI increases the importance of professional responsibility. Technology may generate outputs, but it does not assume accountability. Submitting AI-generated content to the tax authorities without appropriate review creates significant professional and regulatory risks. The ability to critically evaluate AI-generated work is now therefore a fundamental professional requirement.
For international clients establishing a presence in Germany, there is a further consideration. AI-enabled processes must comply with German and European legal requirements, particularly in the areas of data protection, tax compliance and corporate governance. Many foreign investors underestimate this complexity. This is precisely where a firm with both Big Four experience and an international outlook can add significant value: we understand both the technological opportunities and the regulatory realities.
What governance and oversight structures are organisations implementing to ensure AI is used responsibly?
In Germany, AI governance is not simply a matter of best practice anymore; it is becoming a legal requirement. With the EU AI Act, which began its phased implementation in August 2024, a binding regulatory framework is emerging that directly affects tax advisors and financial service providers. AI systems used in regulated sectors will often be classified as high-risk applications, triggering obligations relating to risk assessments, documentation, transparency and demonstrable human oversight.
Professional regulations provide an additional layer of governance. Confidentiality obligations under Section 57(1) of the German Tax Consultancy Act (StBerG) apply equally to digital tools. Client data must not be entered indiscriminately into publicly accessible AI systems and platforms that use submitted information to train or improve models can quickly create compliance concerns. Data processing agreements with AI providers are essential, yet they remain an area that is frequently overlooked.
At Kanzlei Schnürch, client data is processed exclusively within European, GDPR-compliant cloud environments, a standard maintained in close cooperation with our Data Protection Officer. Tools operating outside European jurisdictions are categorically excluded. Decisions regarding the use of AI applications are not made on an ad hoc basis but are subject to a structured review and approval process that considers both data protection requirements and professional confidentiality obligations.
We are currently evaluating purpose-built compliance platforms such as LexTrace, which support organisations in assessing AI use cases against the requirements of the EU AI Act through structured risk classification and audit-ready documentation. Technology can assist the process, but responsibility for decision-making remains firmly with people.
What liability or risk issues arise when professionals rely on AI-generated analysis?
The fundamental principle of German liability law remains unchanged in the age of AI: responsibility rests with the tax advisor or the company’s management, regardless of the tool used to generate a result. An AI system cannot bear professional liability; only an individual can.
In practice, one of the most significant risks stems from so-called AI hallucinations. AI systems can produce highly convincing content that is legally incorrect, based on outdated tax rules or entirely fabricated. Professionals who rely on such outputs without appropriate verification risk submitting inaccurate tax returns, breaching compliance obligations, causing financial losses and, in the most serious cases, jeopardising their professional licence.
Data protection presents an equally important and often underestimated risk. The use of AI tools hosted outside European jurisdictions to process sensitive client information may constitute a clear breach of both the GDPR and professional confidentiality obligations, with potentially serious consequences for internationally active businesses and corporate groups. From a corporate governance perspective, management teams that implement AI without adequate documentation, oversight mechanisms or internal controls expose themselves to increased regulatory scrutiny and potential personal liability.
Our position is therefore straightforward: AI should be used as a carefully selected and closely supervised tool, always subject to human oversight and ultimate accountability. International clients entering the German market do not need a fully automated compliance process. They need a trusted advisor who understands the complexity of German tax and regulatory requirements, communicates clearly in both German and English and remains available when challenges arise. That has been our commitment since 2001 and no algorithm will change it.
