AI in auditing and accounting: A powerful tool that still requires professional judgment.


Artificial Intelligence (AI) has become a common tool within the accounting and auditing profession. Platforms such as ChatGPT and Claude are increasingly used to assist with research, report drafting, summarising information, and improving the quality of written communication. There is no doubt that these tools can save time and improve efficiency in practice.
However, despite the benefits, professionals should remain cautious when using AI in technical fields such as accounting, auditing, tax, and legal interpretation. AI systems are still developing and can produce incorrect answers, misinterpret legislation, or provide conclusions that appear convincing but are not fully accurate or applicable to a specific situation.
A growing concern within the profession is the extent to which some users rely on AI-generated responses without performing independent verification. While AI can be useful as a starting point for research, it should never replace professional judgement or direct reference to the relevant standards, legislation, regulations, and official guidance.
Auditors and accountants have a responsibility to ensure that their work is based on facts, supported by appropriate evidence, and compliant with the applicable laws and professional standards. Information obtained from AI should therefore always be checked against the original source material, including:
Accounting and auditing standards
Tax legislation
Company law and regulations
Regulatory guidance
Industry-specific requirements
One of the risks with modern AI platforms is that they are designed to be conversational and helpful. In some cases, the system may provide responses that align too easily with the user’s assumptions or preferred viewpoint. This can create a false sense of confidence, particularly where complex interpretation or professional judgement is required.
For this reason, users should approach AI outputs with professional scepticism. It is important to critically assess whether the response is reasonable, technically correct, and supported by the relevant legislation or standards.
Where possible, users should also adjust platform settings to improve response quality. Recommendations include:
Prioritising accuracy rather than speed
Requesting detailed explanations instead of brief summaries
Asking for references to relevant standards or legislation
Requesting alternative interpretations or possible risks
Asking the system to identify assumptions or areas of uncertainty
Useful prompts may include:
“Please provide the relevant section of the legislation or standard.”
“What are the alternative interpretations?”
“What factors could change this conclusion?”
“Please provide a conservative interpretation.”
AI can nevertheless be extremely valuable when used appropriately. It is particularly effective in:
Drafting professional reports
Improving grammar and spelling
Structuring documents
Summarising lengthy information
Assisting with administrative tasks
These functions can improve efficiency and help professionals communicate more clearly and professionally. However, the final responsibility for the accuracy and appropriateness of the work still rests with the auditor or accountant.
It is also important to recognise when specialist input is required. AI should not replace consultation with qualified experts in areas where specialised knowledge is necessary. Where uncertainty exists, professionals should still engage with tax specialists, legal advisors, valuation experts, or other relevant professionals to ensure that opinions and conclusions are appropriate and aligned with current legislation and industry practice.
The accounting and auditing profession is built on trust, competence, independence, and professional judgement. While AI is a useful tool that can enhance productivity and save time, it should support professional work rather than replace critical thinking and technical expertise.
In conclusion, AI platforms can provide significant benefits to auditors and accountants, particularly in research and report writing. However, these tools should still be used with caution. Information generated by AI must be independently verified, assessed against the relevant standards and legislation, and supported by sound professional judgement. Ultimately, responsibility for the quality and accuracy of the work remains with the professional, not the technology.
