Guidelines in the Field of Artificial Intelligence (AI) of G & W Steuerberatungs GmbH

G & W Steuerberatungs GmbH places great importance on the responsible use of artificial intelligence (AI). This guideline is intended to ensure that all employees and clients understand how AI is used in our processes and what measures are taken to guarantee security, transparency, and compliance.

G & W Steuerberatungs GmbH uses AI-based applications and services for purposes such as research, improving customer service, inspiration, analysis tools, automation of routine tasks, trend forecasting, customer-specific recommendations, document review, risk assessment, etc.

Employees of the company are permitted to use AI applications and services.

The use of the company email address for registering with AI systems for business purposes is allowed for employees. In this context, the following internal reporting obligations apply: request and approval by the managing directors. Company personnel must comply with all legal and labor obligations as well as these AI guidelines in all activities.

The following examples illustrate how AI tools are applied in our company:

• Automation of routine tasks: Automating repetitive tasks such as receipt capture, data categorization, or preparatory accounting work.
• Optimization of tax strategies: Analyzing tax data to optimize client tax strategies, e.g., by simulating different scenarios.
• Trend forecasting: AI-supported analyses to predict financial trends and potential market developments.
• Compliance with legal requirements: Assisting in the review of compliance in the fields of tax and data protection through continuous analysis of changes in legal texts.
• Client-specific recommendations: Automated, personalized tax and financial recommendations based on individual client data and requirements.
• Risk assessment: Using AI to identify and assess financial risks or risks in business processes.
• Document review: Automatic analysis and review of contracts and other important documents for tax relevance or risks.
• Accounting and tax consulting: Automated systems support accounting processes such as scanning and categorizing receipts. However, decisions on complex tax matters are made exclusively by qualified employees.
• Customer communication: Automated systems assist in responding to common inquiries, while complex issues continue to be handled individually by our advisors.

These use cases help us apply AI purposefully across various areas of our operations. However, no technical tool can ever replace personal contact with our clients!

Our AI systems are operated in accordance with the following principles:

• Accountability: Every AI application in our company is clearly assigned responsibility and properly documented.
• Transparency: Clients are informed when AI is involved in a process that affects their data or decisions.
• Fairness and Non-Discrimination: We ensure that our AI applications do not have discriminatory effects on clients or employees.
• Reliability: AI systems must operate in a stable and predictable manner. Potential sources of error are minimized.
• Data Protection and Security: The protection of personal data is a top priority. This is especially true for sensitive data, which is handled with strict security measures.

AI-generated content is clearly and transparently labeled to ensure that users and affected parties know when and where AI is being used. Our labeling guidelines include:

• Clear indication of AI usage: Content that has been fully or partially created by AI should be marked with a clear note such as “This text was created with the help of Artificial Intelligence (AI).” In customer communication, for example, the note “Automatically generated response with AI support” could be used.

• Identification of AI-generated media: For images, graphics, or videos created by AI, labels such as “AI-generated image” or “This image was created by Artificial Intelligence” should be used.

• Transparency in decision-making: If AI has played a significant role in decision-making (e.g., in the analysis of tax data), this should also be disclosed, e.g., “This result is based on an AI-supported analysis.”

• Automated emails and reports: Automatically generated emails or reports could include a footnote such as “This report was generated using AI software and reviewed by an employee.”

• Visible labeling for chatbots and interactive tools: When customers interact with a chatbot or an AI-supported tool, there should be clear notices such as “You are communicating with an AI-based system” or “This chat is supported by AI.” The labeling is always prominently displayed and easy to understand to ensure transparency and avoid misunderstandings.