Insights > Valuation Operations
What Happens to Professional Judgment When AI Can Do More of the Work?
One of the things I hear consistently when talking with valuation professionals about AI is that human judgment will always be essential to the valuation process.
And I agree with that. At least in principle.
But I think we need to examine what we actually mean by judgment, because I’m not sure we’re all talking about the same thing.
For years, professional judgment has been one of the defining characteristics of valuation expertise. It’s what separates an experienced valuator from someone who simply knows how to build a financial model or apply a methodology.
The assumption has generally been that while technology can automate calculations and repetitive tasks, judgment is something that remains firmly in the hands of the professional.
I’m beginning to question how much of that assumption will hold.
Not because I believe AI will eliminate the need for valuation professionals. But because some of the activities we’ve traditionally classified as professional judgment are becoming increasingly possible for AI to perform.
And I think that distinction matters.
What Do We Actually Mean by Judgment?
Consider the different activities that go into a typical valuation engagement.
We analyze financial information, identify anomalies, evaluate assumptions, select methodologies, assess risk, interpret results, and ultimately arrive at a conclusion of value.
Throughout that process, we exercise what we call professional judgment.
But these aren’t all the same kind of activity.
Some involve pattern recognition. Others involve applying established analytical frameworks, comparing alternatives, or identifying inconsistencies in financial information.
Those are areas where AI capabilities are advancing rapidly.
I’ve spent considerable time recently testing newer AI models against actual valuation work. Not simply asking them to summarize documents or generate text, but having them construct sophisticated financial models with multiple layers of calculations and interdependent formulas.
The results have been significant.
In one test, I provided instructions for a complex valuation spreadsheet that would ordinarily require substantial analyst time to construct. The model generated a structurally sound spreadsheet in minutes.
Now, was it a finished valuation? Of course not.
There were still assumptions to evaluate, calculations to verify, and conclusions that required professional review.
But the technical execution was much further along than I would have expected even a short time ago.
That experience raises a larger question.
If AI can perform increasingly complex analytical work, how much of what we’ve historically described as judgment is actually a collection of activities that can be separated, evaluated, and potentially automated?
Not All Judgment Is the Same
I think it’s useful to distinguish between several different components of professional judgment.
Analytical judgment involves recognizing patterns, identifying unusual relationships, evaluating financial performance, and detecting inconsistencies.
Methodological judgment involves determining which valuation approaches are appropriate, selecting relevant inputs, and evaluating alternative assumptions.
Contextual judgment involves understanding the circumstances surrounding an engagement, including the purpose of the valuation, the characteristics of the business, and factors that may not be fully captured in the financial information.
Professional judgment and accountability involve taking responsibility for the analysis, evaluating whether the conclusions are reasonable, and standing behind the work.
These activities overlap, but they’re not interchangeable.
AI is already demonstrating capabilities in parts of the first two categories. It can identify patterns, perform calculations, compare alternatives, and apply analytical frameworks.
That doesn’t mean it does those things reliably in every situation. It certainly doesn’t mean the results should be accepted without review.
But we shouldn’t confuse a current need for oversight with a permanent technological limitation.
The capabilities are changing, and they’re changing quickly.
The more interesting question is how those changes will affect the responsibilities of the valuation professional.
The Value of the Professional May Be Changing
When clients engage a valuation professional, what are they actually buying?
Historically, we’ve described the answer in terms of technical expertise, experience, and professional judgment.
And those things are important.
But clients are also buying independence, credibility, defensibility, and accountability.
They need a conclusion that can withstand scrutiny. They need to know that the underlying analysis has been properly evaluated, that the assumptions are supportable, and that someone with the appropriate qualifications is willing to take responsibility for the work.
I think those distinctions become more important as AI capabilities improve.
If sophisticated financial analysis becomes faster and more accessible, the ability to produce that analysis may no longer provide the same competitive advantage it once did.
That doesn’t eliminate the value of the professional.
It may change where that value is concentrated.
The professional’s role could increasingly involve directing the analytical process, evaluating outputs, challenging assumptions, resolving inconsistencies, and determining whether a conclusion is supportable.
In other words, less time may be spent producing the underlying work and more time evaluating its reliability and implications.
That is a significant change in how we’ve traditionally organized valuation engagements.
We May Need to Rethink How Valuation Work Is Performed
There’s another implication that I think deserves attention.
Most valuation firms have developed their processes around the capabilities and limitations of the people performing the work.
Analysts collect and organize information, build models, perform calculations, and prepare initial drafts. More experienced professionals review that work, evaluate assumptions, and make decisions about the conclusions.
That structure has developed over many years, and for good reason.
But what happens when some of those activities can be completed in a fraction of the time?
One response is simply to introduce AI into the existing process and make each step more efficient.
That’s useful, and there are certainly opportunities to improve productivity that way.
But I think there’s a ceiling on the benefits of that approach.
The larger opportunity is to reconsider how the work should be performed in the first place.
For example, if AI can generate a sophisticated financial model, should we continue organizing the engagement primarily around the time required to build that model?
Or should we reconsider the process so that more attention is devoted to evaluating the assumptions, testing the results, and documenting the basis for the conclusions?
The same question applies to financial analysis, report preparation, and quality control.
We shouldn’t assume that every step in today’s valuation workflow needs to exist in its current form simply because that’s how we’ve always done it.
At the same time, we need to be careful not to eliminate necessary controls in the pursuit of efficiency.
AI-generated work still needs appropriate validation, and professional responsibility doesn’t disappear because a system performed the underlying calculations.
The challenge is determining what the right combination of technology, professional oversight, and documented review should look like.
Accountability May Be the More Durable Differentiator
I don’t think the valuation profession is going away.
I do think the nature of the work, and potentially the way we define expertise, will change.
For a long time, a significant part of the value we provided was tied to our ability to perform complex technical work that clients couldn’t easily do themselves.
As those capabilities become more accessible, firms will need to think carefully about what continues to differentiate them.
My view is that professional accountability, independence, credibility, and the ability to defend a conclusion will become increasingly important.
AI can support the analysis. It can help identify issues, evaluate alternatives, and produce sophisticated work product.
But generating an output is not the same thing as accepting professional responsibility for a valuation conclusion.
That distinction matters to clients, regulators, courts, and anyone relying on the work.
And it should matter to valuation firms as they determine how to incorporate these capabilities into their practices.
A Different Conversation About the Future of Valuation
I suspect we’ll continue hearing that AI can automate technical work but that human judgment will always be necessary.
I don’t necessarily disagree.
I just think that’s too broad a statement to be particularly useful.
We need to start identifying which elements of judgment are becoming more automatable, which require a different form of professional oversight, and which responsibilities must remain with the credentialed professional.
That has implications for how we train analysts, structure engagements, develop review procedures, and ultimately define the value we provide to clients.
And those questions are becoming increasingly relevant as the technology improves.
Rather than simply defending the importance of human judgment, I think the profession would be better served by examining exactly what that judgment consists of—and how its role may change.
That’s a much more productive conversation to be having.
Better valuation operations begin with better questions.
Explore how ValueVision is building specialized agents, connected operating infrastructure, and practical advisory services for the future of valuation.