The IEAI Discusses AI Ethics Training for Accountants

The IEAI Discusses AI Ethics Training for Accountants

“AI Ethics at Work: Reflections on Designing Training for Accountants”

As AI becomes embedded in professional work, the question is not whether existing ethical principles still apply but how their application must change.

How can we train professionals to use AI responsibly, particularly when their knowledge and experience of it may be limited? Creating ethics training for such a diverse audience presents its own set of challenges.

Invited speaker, Emeritus Professor Christopher Cowton of the University of Huddersfield, addressed this question at the IEAI June 2026 Speaker Series. Prof. Cowton, a prominent figure in business and financial ethics and a Fellow of the Academy of Social Sciences, shared his experience creating an online course on AI ethics for the Institute of Chartered Accountants in England and Wales (ICAEW), one of the top professional accounting organizations in the UK with about 140,000 members.

The task was far from straightforward. When Prof. Cowton was approached, his most recent reading on AI dated back to Joseph Weizenbaum’s early work on the ELIZA program. He illustrated his starting position with Bruegel’s painting The Blind Leading the Blind: an author still learning about AI, writing for professionals whose own knowledge of it varied widely. The challenge was compounded by the field’s rapid pace, its constantly changing terminology and the scarcity of existing material on AI ethics for accountants. The audience itself was very diverse, from people working for the Big Four to small local firms and accountants working in business. Prof. Cowton prepared by reading books and online materials and by running four roundtables that brought together professional accountants and AI experts. The aim was not to become an AI expert himself but to learn enough to frame the ethical discussion.

The resulting module, which was released in January 2025 as part of ICAEW’s broader ethics course for ongoing professional development, deliberately keeps its introduction to AI brief and instead focuses on ethical risks such as bias, hallucinations, sensitive data, security and intellectual property. Prof. Cowton stressed that many of these risks call for competent practice as much as philosophical reflection.

There are still ethical issues, yes, but an awful lot of them can be mitigated, if not eliminated, by just good AI practice.

The module revisits the five core principles of the ICAEW Code of Ethics and demonstrates how the existing framework can be applied to AI scenarios through a case study of an AI fraud detection system. This echoes the talk’s key point: AI ethics should not be considered a world apart, separate from the professional ethics accountants are familiar with.

You’re still in the ethics world, or at least you’ve got one foot in the ethics world.

By April 2026, nearly 19,000 members had completed the module, and the feedback was overwhelmingly positive. Looking back on the experience, Prof. Cowton commented that while the widely circulated principles of AI ethics seem relatively stable, discussions about AI ethics do not yet appear to be maturing or deepening. During the lively Q&A that followed, the discussion turned to whether the profession needs a new AI-specific principle (in his view, existing principles are abstract enough, though their application must be updated for the AI context), how delegating work to AI compares to delegating to junior colleagues and whether there may even be an ethical risk in not using AI for tasks where it can perform better.

We extend our appreciation to Prof. Cowton for sharing his valuable insights and for engaging with the IEAI community on this timely topic. We would also like to thank all our event registrants for their participation and the TUM Think Tank for hosting. The event recording
can be found below.