(Credit: Thapana Studio - adobe.stock.com)

In a review published on Thursday, the watchdog noted that while the Big Four firms – Deloitte, EY, KPMG, and PwC – alongside BDO and Forvis Mazars, increasingly use automated tools for risk assessments and gathering evidence, they fail to track the impact on audit quality. The FRC stated there was “no formal monitoring performed by the firms to quantify the audit quality impact”. Instead, monitoring has primarily focused on tracking usage for licensing purposes.

The review was published alongside the FRC’s first guidance on the use of AI in audit.

The FRC’s review, prompted by its audit quality team, found that all but one of the firms had no key performance indicators for these tools. This comes as firms are heavily investing in AI to improve efficiency. Deloitte, for example, uses AI to summarise board minutes and complex contracts, while KPMG employs it to scan millions of transactions to identify risks — a task not possible with traditional methods.

While the FRC acknowledges AI’s potential to improve audit quality, it also highlights “risks and challenges”, including ethical issues and potential bias in the technology’s output. 

Mark Babington, FRC executive director of regulatory standards, urges firms to define metrics to evaluate their impact.

He said: “AI tools are now moving beyond experimentation to becoming a reality in certain audit scenarios. When deployed responsibly, they have significant potential to enhance audit quality, support market confidence, drive innovation and ultimately contribute to UK economic growth.

“The FRC continues to support and encourage innovation in audit. This guidance aims to illustrate how AI can enhance audit work as well as clarify FRC expectations around proportionate, appropriate documentation of tools that use AI.

“We recognise that this field is moving quickly and will continue to engage across the profession, both in the UK and internationally, to support innovation and the appropriate use of AI.”

Key features of the FRC’s guidance:

• Two-part structure – Illustrative example of one potential way AI can be leveraged in an audit, as well as principles that are intended to support proportionate and robust documentation of tools that use AI
• Broad and forward-looking AI definition – Encompasses both traditional machine learning and deep learning models, including generative AI
• Balanced documentation expectations – Proportionate approach to prevent over documentation
• Sophisticated view on appropriate explainability – Acknowledges that appropriate levels of explainability vary based on context and usage
• Versatile principles – Illustrates topics, judgements and considerations that have broad applicability to other instances of AI use in audit
• Alignment with Government AI principles - Documentation guidance reflects the UK government’s five AI principles
• Relevant across market – The guidance contains material that clarifies how expectations translate into contexts where a tool is obtained from a third party