2023: How artificial intelligence helps audit
Artificial intelligence algorithms can improve audit efficiency. In particular, such tools help identify cases of fraud. This is stated in a study by the international audit consulting corporation EY, the results of which were disclosed in early December 2023.
According to Kath Barrow, managing partner of EY in Britain and Ireland, an AI system was used to detect suspicious activity in two out of 10 tested companies. These customers subsequently confirmed they had faced fraud.
EY argues that the AI-based technology developed is more complex than widely used software designed to detect suspicious transactions. The new system is trained using information about a wide variety of fraud schemes: the data is taken from both open sources and from investigative materials conducted by EY itself. AI algorithms are able to identify not only suspicious transactions, but also operations that are usually used to hide fraud. If questionable actions are detected, the system issues an appropriate warning, which allows auditors to conduct additional analysis.
It is noted that AI can potentially significantly reduce the burden on the specialists of audit companies. Such tools are able to analyze huge amounts of data and improve themselves over time. At the same time, AI is designed to assist auditors, and not replace them.
Some experts, however, are skeptical that AI can work effectively during an audit. So, KPMG says that fraud is inherently unpredictable, and therefore the use of previously accumulated data for training AI models may not give the desired result: the neural network simply will not be able to identify new criminal schemes.[1]