The adoption of advanced language models, such as ChatGPT, in the accounting field has sparked considerable interest and discussion in recent times. In this blog post, we dive into a recent paper that talks about how Uniper, a multinational energy company, has harnessed ChatGPT to enhance its internal audit function (IAF) and revolutionize its auditing processes. By adopting ChatGPT, Uniper has achieved remarkable efficiency gains ranging from 50 to 80 percent in various audit processes. This real-world example will provide valuable insights for practitioners seeking to leverage ChatGPT for auditing and researchers interested in the application of large language models in the corporate domain.
In late 2022, the advent of advanced language models sparked a wave of interest across various industries. These models, such as ChatGPT, have demonstrated their ability to recreate knowledge and perform simple tasks, making them accessible through low-code interfaces. This blog post specifically explores how Uniper, a multinational energy company, has taken the initiative to incorporate ChatGPT into its internal audit tasks, revolutionizing the way audits are conducted. By employing ChatGPT, Uniper’s internal auditors have witnessed significant enhancements across multiple phases of the auditing process, from risk-based audit planning and preparation to the actual audit procedures and data analysis.
The internal audit function plays a crucial role in a company’s internal control, risk management, and governance functions. However, internal audit teams often face challenges related to budget constraints and effectiveness. The implementation of technology, particularly large language models like ChatGPT, offers the promise of improving efficiency and effectiveness despite limited budgets. Therefore, exploring how ChatGPT can facilitate the construction of audit steps, extract insights, draft working papers, improve report readability, and streamline data analysis tasks is of utmost importance.
Language models like ChatGPT have earned recognition for their ability to generate coherent and contextually appropriate text responses, making them an intriguing tool for auditors. Previous research has proposed the concept of copiloting, which involves a collaborative interaction between humans and AI systems to complete tasks. In this approach, auditors maintain control over the direction and procedures of the audit while utilizing the language model as an audit-oriented assistant, providing valuable support, insights, and analysis to aid decision-making.
To put copiloted auditing into practice, the paper examines how Uniper is using ChatGPT as a copilot to enhance the efficiency and effectiveness of its internal audit processes.
Find out more here.
