The Internal Control System and AI Applications

The applications of artificial intelligence (AI) are diverse and must be chosen with care. In addition to technical considerations, practical safety requirements must be taken into account.

Through the use of artificial intelligence, state-of-the-art forms of automation are set to unlock new potential for efficiency. Whilst simple automation solutions, such as Robotic Process Automation, have been available on the market since the early 2000s, machine learning (ML) and artificial intelligence offer far-reaching opportunities and potential for optimisation. In particular, the ‘self-learning’ capability opens up a wealth of application areas for improving the quality of decision-making in previously manual tasks, even across different interfaces. The use of AI in finance and accounting processes therefore promises not only the automation of repetitive process steps but, unlike Robotic Process Automation, also the ability to take on qualitative decisions that previously had to be made by humans.

Volatility: new challenges for process monitoring

The extensive and often complex capabilities of AI, particularly its ability to learn autonomously, also place greater demands on companies’ corporate governance frameworks. This is because, in order to achieve results of sufficient quality using machine learning and artificial intelligence, AI-based algorithms must constantly learn, evolve and adapt. This volatility poses new challenges for the management and monitoring of processes and IT systems.

Furthermore, the quality of the data used can lead to incorrect decisions and breaches of regulations. For example, AI algorithms used to assess credit applications from bank customers may be programmed in a discriminatory manner or set up using biased test data. As a result, certain customer groups (age groups, ethnic origins, etc.) could be treated in a discriminatory manner. This demonstrates that the risks associated with AI are not merely financial in nature,

Ready to talk? Contact us

auto_stories

Our corporate governance update. This time in focus: Artificial intelligence between efficiency and responsibility

Download now

How KPMG supports businesses

New quality assurance processes and controls must be implemented to ensure the stability of AI systems and the traceability of the decisions they make.

KPMG supports companies in the efficient and secure use of artificial intelligence by:

Design and implementation of process and control structures in both the IT and operational areas, based on the KPMG AI in Control framework
Design and implementation of an AI-focused governance structure incorporating elements such as, for example,
- Inventory and risk assessment
- Training and communication measures
- Building trust (AI Code of Conduct, expanded guidelines, system-level controls, quality standards, etc.)
Identifying opportunities for automation using artificial intelligence within your processes and for tasks that have previously been carried out manually
Review of the governance structure and internal control system relating to AI systems, the AI-specific operating model, and individual algorithms relevant to financial reporting (relevance in the context of the statutory audit).

The KPMG AI Governance Team brings together data scientists and process and governance experts with extensive knowledge of auditing and advisory services in the field of corporate governance, with a focus on the use of digital solutions. We therefore provide support in establishing and operating a targeted AI environment that ensures secure and efficient AI operations and enables the full potential of automation solutions to be realised.

Interested in our services? Contact us