The future is now. And audit is back.

Not too many years ago, serious people told me auditing was a dying profession. Commoditized, stagnant, irrelevant. I disagreed. In fact, after spending more than 40 years in the profession, I believe that this is the most exciting time in the history of the audit profession. And the driving force is principally the transformational power of AI.

Today, the same forces that some thought would hurt audit (technology, automation, AI) have made it more essential, more complex, and more consequential than ever. The capital markets, the regulators, the boards of directors who depend on independent attestation, all need a robust audit profession now, more than ever.

But being essential doesn't mean being safe. Ten years from now, the profession will look nothing like the one that exists today. The firms that understand the exponential speed of transformational change that is empowered by AI are sprinting ahead. The ones that don't are at risk.
the station

‍

AI is a total business transformation, not a technology upgrade

Several years ago, I stood in front of a group of global business leaders and said "transform or die." Admittedly, that was a bold statement and it was intentionally provocative.  The profession was not in danger of "dying" in the traditional sense.

My point was this: the speed of AI-enabled change will require audit firms to embrace comprehensive transformation, or risk falling into a death spiral.

What is this so-called death spiral? I believe that firms that do not embrace the change and transform likely will see their operating margins erode because they will not be capturing and managing their entire portfolio of costs. When margins erode, pressures to cut corners increases. Cut corners in audit and quality drops, failures rise. Audit is too important to the public interest for this to happen. That is why I implored the audience to embrace the change.

To be clear, AI enabled technology is driving the transformation of the audit profession. But one cannot address such technology in isolation. One must embrace the multiple dimensions of its impact on the audit business and manage change across all of them simultaneously.

Let me give you a simple anecdote to illustrate this point. In speaking to an audit committee chair on the emerging power of AI in the audit, I was asked only one question: 

“Larry, how many hours will this save and what is the reduction of our fee?” 

I was taken aback. I had been extolling the virtues of a more effective audit. But I could not blame him. For more than one hundred years, the profession ran on a "rate × hours" basis and that is also how we essentially communicated to those charged with governance. Nonetheless, his question missed the point - and, in self reflection, I did not adequately communicate the transformational impact that I expected from the power of AI and why.

As a result, I want to communicate to interested parties that the impact of AI on the audit  profession is truly a multi-faceted issue. AI doesn't touch one part of the audit business. It touches all of it. Draw a circle with AI at the center and consider the various aspects of the operations of an audit practice that sits around the edge:

• Recruitment: what skills do we actually need from a university graduate now?
• Training: what changes need to be made to our continuing education programs?
• Career paths: is the “traditional” career path of moving from audit staff to manager to partner still the right model, or is the career path more irregular?
• Pricing: are we still billing on a "rate × hours" basis when AI compresses certain tasks from a week of work into an afternoon?
• Cost structures: are all technology costs being accurately captured, including cybersecurity, cloud, product research and development, and the rapid pace of technological obsolescence?
• Delivery models: has our use of delivery centers become obsolete? Should we develop Centers of Excellence?
• Communications with regulators and to those charged with governance: do we still communicate with audit regulators and with Audit Committees under our old model and have we adequately explained how AI is being used in the audit?
• Risk assessment: have we recognized that the companies we audit are also using AI in their own financial reporting processes and has the audit approach been accordingly? As an example, what about the increased risk of fraud?
• Methodology: have we adequately adjusted our audit methodology to keep pace with these changes?
• Technology: have we fully integrated AI into our delivery tools and our audit methodology or is it merely a “bolt-on” product?
• System of Quality Management:  have we adapted the controls in our firm-wide system of quality management to address a new set of risks?

I think you get the point by now! Every component in the audit process and in the management of an audit business has the potential of being transformed. A firm that exclusively upgrades its technology stack but leaves everything else untouched hasn't transformed. It's merely rearranged the furniture.

In a series of future posts, I will examine the components of the audit business most affected by this transformation and the catalyst behind it: the exponential pace of AI. But before we unpack each of those changes, let's start with the bigger picture.

‍

The historical delivery model is rapidly becoming obsolete

The leverage pyramid (say for example, one partner for every 10 to 12 professionals) was designed for a manually intensive audit process where seasoned experience and insight bottlenecked at the top. That logic is rapidly changing.

What's replacing it:

• Fewer junior staff are entering the business
• Delivery centers are being used in fundamentally different ways, not simply as cost reduction mechanisms in low-cost jurisdictions, but as centers of coordinated, technology-enabled execution
• Subject matter experts and Centers of Excellence are being elevated, because business complexity now demands genuine depth, not generalists wearing specialist hats

The firms navigating this well aren't just restructuring organization charts. They are rethinking what an engagement team is, who is on it, and what each function is accountable for.

‍

The "rate x hour" model must change

For most of the profession's history, a "rate × hours" model made sense. Effort, progress towards completion of the audit engagement and value were correlated. That correlation is dissolving, and the pressure doesn't stop there, because new or incremental costs arise in numerous other areas.

Firms that keep billing by the hour as AI reduces hours expended will watch their margins decline, while the costs of technology development, deployment, training, cybersecurity, and infrastructure climb at an increasing rate. You cannot capture the efficiency benefits of AI while pricing as if nothing has changed. In many cases, those other technology related costs are accounted for as “overhead” rather than direct costs of the delivery of the audit engagement.

The shift the profession must make? Value- and outcome-based pricing: pricing fundamentally built on client complexity and the value of the assurance delivered, not hours logged.

Other professional services businesses have  already made this move. Firms that get this right will decouple revenue from headcount and earn revenue for the value and audit quality they deliver. How do firms manage this change? I will give you my view in a future post.

This shift in approach to pricing and running the audit business is the biggest change management challenge in the profession's history. Whether firms seize it or let it pass will define the next decade.

‍

The "black box" problem nobody has solved

Here is the question that must be addressed by every audit firm utilizing AI: how do you know if an LLM's output is right, to a standard that constitutes defensible audit evidence?

Regulators don't want a black box. They want an auditor who can open it, walk them through what's inside, and demonstrate that a rigorous system of quality management (back testing, paradigm testing, documented controls) governs every step.

"I put the data in and this is what came out" is not an answer.

I do believe that the audit profession recognizes this issue and is working to address it.  But, the profession has not yet converged on a single set of metrics, right confidence thresholds, or a consistent vocabulary for what "sufficient" means when the output comes from a model rather than a human.

International Standard on Quality Management No. 1 (ISQM 1) issued by the International Auditing and Assurance Standards Board (IAASB) provides the framework and the obligation to address key aspects of this issue. This is a defining technical and regulatory challenge of the next five years. The firms that crack it first will have a durable competitive advantage in quality and relevance.
‍

Why I'm supporting Cortea

The make versus buy debate has effectively been settled. AI development is exponentially fast, and audit firms that try to build everything internally are already falling behind those that leverage the speed and innovation of third party developers.

Large platform partnerships are a step forward. But they can move slowly, and they tend to optimize for the broad market rather than the specific, high stakes requirements of audit.

Lean, focused startups building products specifically for audit move faster. They understand the regulatory constraints from day one. And critically, they can solve the hardest problem: making AI output reviewable and defensible in front of a regulator.

That's why I'm supporting Cortea. It's not about making auditors faster. It is about the core accountability of quality and making the output explainable and auditable. The firms that partner aggressively with the right startups will outpace the ones waiting for their internal platform to catch up.

The firms that move now, even imperfectly, will own that future. The ones waiting for certainty will find the market already divided by the time they decide to act.

‍

The train has left the station

Two years ago I said publicly that AI needed to be on the agenda of every audit committee meeting, starting immediately. The sme went for every meeting between the audit profession, standard setters, clients, and regulators. Those in the room nodded. Then they went back to their playbooks.

My strong preference, then and now, is to go on this journey together: regulators, firms, and the profession.

But the journey is happening regardless. The only question is who shapes it.

The profession isn't dying. It never was. It is being remade at a pace none of us fully anticipated. I believe in the audit profession. The future is now. And audit is back.

‍