Artificial intelligence (AI) broadly refers to technologies that make machines “smart.” AI continues the trend in advancement of technology that has unleashed the practical applications, many of which can enhance the decision-making process. AI is powered by algorithms, and algorithms are driven by large amounts of data.
Auditing is an essential function for organizations, but much of it is routine. It lends itself to the use of technology to analyze big data and decide which areas of the audit to focus upon and how best to gather the data needed to ensure the audit meets professional and ethics standards. Accounting firms are experimenting with AI, in which machines go beyond doing rote tasks and inform basic decision-making.
There are many implications of AI for the corporate governance system including internal auditing, internal control over financial reporting, and the role of the external auditors. Moreover, unintended consequences may exist with respect to the potential for fraud in AI systems and developing efficient and effective AI systems.
What’s missing from today’s discussion of AI in the accounting and auditing arena is identification of specific issues of concern as AI systems evolve. These include the following.
How can an organization establish accountability and oversight through corporate governance systems in an AI environment?
How does the use of an AI system influence the role and responsibilities of the Chief Financial Officer? This can be linked to the requirement to certify financial statements as not containing any material misstatements, as required under Section 302 of the Sarbanes-Oxley Act.
Can AI systems be used to promote the agenda of management that may include occupational and/or fraudulent financial statements. If yes, how can the integrity of the financial statements be protected?
What competencies will need to be developed to enable internal auditors to promote AI-related advisory and assurance services?
How can AI systems be used to facilitate specific audit routines – i.e., accounts receivable confirmations?
How might AI change the composition of the audit teams in CPA firms?
What is the role and responsibilities of the system of internal controls over financial reporting in an AI environment? This can be linked to management’s requirement to assess whether internal controls in an AI system are operating as intended and the external auditor’s role in reviewing management’s report and coming to an independent assessment, as required under Section 404 of Sarbanes-Oxley.
What are the role and responsibilities of the audit committee in an AI environment?
What are the possible threats to objectivity and integrity in an AI environment and the safeguards to mitigate those threats and enhance the external audit function?
What role might the fraud triangle play in preventing and detecting fraud in an AI environment? This could be done by identifying the internal pressures under AI to manipulate financial data, having the opportunity to do so, and rationalizations after the fact as to why it was done?
What are some of the societal/ethical issues of using AI systems.
Will robots take over the job of accountants?
What specialized training is necessary in the audit environment?
What are the implications for data privacy and the ethical use of personal data?
What are the implications of using AI systems to conduct the audit with respect to the need to exercise a degree of professional skepticism in gathering useful data for the audit?
The challenges for accounting, auditing, and financial reporting in an AI environment are substantial. One way to characterize it is to emphasize the need for competency, diligence, and ethical practices. Another is to “trust but verify.”
Steven Mintz is a writer, speaker, and consultant on ethics issues. You can contact him at firstname.lastname@example.org.