Tips for accountants on the ethical use of AI

Artificial intelligence (AI) is developing at lightning speed, with more accountants using it for everyday tasks – from invoice processing to data analysis to forecasting. Despite the productivity benefits, AI poses ethical risks accountants would do well to consider.

by | 31 Jan, 2025

Accountants have embraced AI for everyday tasks such as data analysis, but this can pose ethical risks. 

This expansion can increase efficiency and reduce costs. However, it raises a range of ethical issues. 

How could AI threaten clients’ privacy? What kinds of biases inform AI-driven findings? What happens when AI makes errors? 

Despite these and other concerns, there is no regulatory framework governing AI in the UK – though existent laws cover some aspects of its use. For example, the General Data Protection Regulations (GDPR), limits personal data collection and restricts automatic decision-making where it could have a major impact on someone’s life.

So, how can accountants ensure that their approach to AI is ethical? We ask Dr Mustafa Sayim, Senior Lecturer, London School of Business and Finance, and Ian Gregory, CTO, Advancetrack.

What is ethical AI?

“Ethical AI entails designing and deploying artificial intelligence systems that prioritise fairness, transparency, accountability and respect for human values,” says Sayim.

Headshot of Dr Mustafa Sayim
Dr Mustafa Sayim, Senior Lecturer, London School of Business and Finance

“The goal is to prevent harm and promote responsible use […] In essence, [ethical AI] assures that AI follows ethical norms while avoiding bias and limiting potential damage.” 

Ethical AI aims to complement human expertise rather than replace it, says Gregory.

“It means that, as accountants, we’re using the technology as a tool to enhance our decision making and efficiency, all the while safeguarding client data and adhering to the strictest professional standards.”

What ethical concerns should accountants consider when using AI? 

The use of AI raises a variety of ethical concerns – for both accountants and their clients.

When it comes to clients, one of the most significant concerns is algorithmic bias, says Sayim.

“[This occurs when] AI models trained on biased data produce unfair results.”

This bias can be difficult to identify, assess and address because AI decision-making processes are often complex, and lack transparency and accountability. 

In the worst case scenario, AI can make significant errors. 

“AI isn’t infallible and can produce outputs that are misleading and incorrect, often known as ‘hallucinations’,” says Gregory. 

“These inaccuracies can have severe consequences if relied upon without human oversight.” 

Headshot of Ian Gregory
Ian Gregory, CTO, Advancetrack

Compounding these concerns are risks to data privacy and security, given that AI often handles sensitive financial data, says Sayim.

Turning to accountants – and the industry more broadly – AI presents the risk of the commoditisation of expertise, which may put pressure on firms to lower fees.

“This could lead to a ‘race to the bottom’ on pricing, where clients undervalue the strategic and advisory expertise of accountants as they see routine tasks being automated,” says Gregory. 

How can accountants ensure they address ethical concerns when using AI?

The first step is ensuring that all AI complies with relevant existent laws and industry standards, says Sayim.

Not only will this help accountants use AI ethically, it will also help avoid legal issues. 

The second step is developing business-specific guidelines.

“Establish a clear and consistent ethical framework for AI usage in your organisation to guarantee that AI applications are in line with business principles, and foster fairness, accountability, and transparency,” says Sayim. 

Next, to reduce the likelihood of bias and errors, accountants should keep a close eye on the results produced by AI.

“It is essential that accountants deploy their critical thinking skills to quality assure the outputs of AI-driven systems,” says Gregory. 

This requires training staff in evaluating outputs – to ask questions such as “Do [the outputs] make sense? Do they reconcile with facts available elsewhere?” 

To prevent privacy and security issues, strict controls should be in place. 

“[These should] avoid sensitive information leaking beyond its intended boundaries, and remove sensitive data from training sets when permission to use it is revoked,” says Gregory.

“The goal is to prevent harm and promote responsible use […] In essence, [ethical AI] assures that AI follows ethical norms while avoiding bias and limiting potential damage.”

Dr Mustafa Sayim, Senior Lecturer, London School of Business and Finance

How can ethical AI benefit accountants? 

Ethical AI isn’t only about doing the right thing. 

It’s also good for business. 

Accountants who establish, and adhere to, ethical standards are more likely to inspire trust in their clients – and trustworthiness is key to expanding business. 

They are in a better position to reap the benefits of using AI.

Perhaps the most significant benefit is increased efficiency. 

“With over 40% of partners reporting they work an extra day a week just to keep up, AI can help by automating repetitive, time-consuming tasks and supporting routine tasks like creating meeting minutes,” says Gregory. 

This, in turn, might help to address staff shortages.

“This allows staff to focus on higher-value analysis work and client interactions, enhancing productivity and making the profession more attractive,” says Gregory.

In addition, as long as AI is monitored carefully for bias and errors, it can help with accuracy. 

“[AI] improves accuracy by evaluating massive datasets to detect errors or abnormalities, resulting in more dependable financial reports,” says Sayim. 

“[It also] dramatically improves company endurance through predictive analytics, which analyses large amounts of data from a variety of sources, including internal systems, global events, market trends, and environmental factors, to construct models that foresee future disturbances.”

Gregory adds, “AI has real potential in surfacing anomalies in data. 

“This will improve assurance levels in audit, help drive fraud detection controls, and drive deeper insights in management accounts. 

“Once again, these will only come to fruition when supported by accountants with first-rate critical thinking skills.”


Want to deepen your knowledge around AI? The IFA will hold a conference on AI on 6 March. More information here.

Share This