The first thing you notice when you walk into a Big Four office on a Tuesday night in midtown Manhattan is how few people are still going through binders. The workstations are more tidy. The screens are larger. The auditor is posing questions to what appears to be a chat window on someone’s screen next to a general ledger, much like you would to a slightly distracted coworker. This is the part where the profession was not forewarned that the change would come in the form of a sidebar.
Auditing has rewarded a certain level of patience for the majority of its modern history. the readiness to settle on a figure. the self-control to double-check. Speaking with senior partners these days gives the impression that they are unsure of how to handle associates who have never had to complete that task the slow way. The muscle memory might still be important. Perhaps it doesn’t.
| Topic Focus | AI’s impact on the U.S. audit and accounting profession |
| Profession Affected | Certified Public Accountants, internal auditors, external auditors, FP&A professionals |
| Estimated U.S. Accountants | Roughly 1.4 million practicing accountants and auditors |
| Key Stat | 72% of S&P 500 companies referenced AI in their most recent 10-K filings |
| Major Regulatory Body | Public Company Accounting Oversight Board (PCAOB) |
| Industry Group | Center for Audit Quality |
| Training Provider Mentioned | Institute of Management Accountants (IMA) |
| Skills Now in Demand | Data literacy, prompt engineering, model evaluation, AI ethics |
| Emerging Assurance Areas | Cybersecurity, ESG, AI governance |
| Cybersecurity Board Expertise (S&P 500) | 60% of boards disclosed a cybersecurity expert in 2024, up from 51% in 2023 |
| Geographic Scope | All 50 U.S. states; impact heaviest in NY, IL, TX, CA |
| Year of Major Shift | 2023–2026 (generative AI adoption phase) |
The figures speak for themselves. Artificial intelligence is now mentioned in the annual filings of 72% of S&P 500 companies. This means that audit teams are required to assess AI-related disclosures regardless of their comprehension of the underlying models. That is a significant change. Sixty percent of S&P 500 boards now have a designated cyber expert, whereas ten years ago, the same companies hardly mentioned cybersecurity in their 10-Ks. It’s a well-known pattern: assurance comes after risk, and risk comes after technology, usually with an uncomfortable lag.
The shift in company culture is more difficult to document. In cities like Chicago and Dallas, younger accountants are coming in with experience in prompt engineering and a casual command of tools that their managers learned from a webinar approximately six months ago. A silent inversion is taking place. The judgment, the connections, and the opinion’s signature are still owned by the elders. However, the child who has been out of school for two years is the one who understands how to get the model to behave.
Beneath the official optimism, it’s difficult to ignore the anxiety. Thoughtful papers about AI enhancing rather than replacing auditors are published by firms. Micro-credentials are being sold by the IMA. Guidelines are being updated by the AICPA. However, you’ll hear a manager questioning out loud on the train ride home whether the company actually needs the same number of employees for the upcoming busy season. The honest response that most people won’t give at work is probably not.
When an algorithm can read a million journal entries in the time it takes a human to read one, the deeper question is what audit actually means. Time and tick marks used to be the foundation of trust. It now needs to be based on something more difficult to define: the auditor’s assessment of whether the machine performed correctly, whether the company’s machine performed correctly, and whether anyone present truly comprehends either machine. The signature is still desired by investors. The name on the opinion is what they want. Human accountability for non-human labor is what they are increasingly paying for.
As this develops, the profession appears to be in the midst of something it hasn’t yet given a name to. The AICPA exam is evolving. The training is evolving. When a model completes what used to take eight hours in eight seconds, the billable hour—that antiquated, dependable concept—begins to seem strange. That contains tension and, although it’s unclear how it will manifest, opportunity as well. What is evident is that every American accountant, from a solo practitioner in Boise to a senior partner in New York, is currently working in a field that is constantly changing. You suspect that those who don’t wait for things to settle will fare the best.
