A practical breakdown of how artificial intelligence is changing the work of appraisers — and why it won’t replace the expert, but can make them significantly more effective.
According to McKinsey, around 70% of companies that have integrated AI into their business processes report EBITDA growth of up to 5%. The technology is already being applied across finance, logistics, and risk management — and increasingly in analytics. A natural extension of this trend is business and asset valuation. But can you really trust an algorithm with calculating the value of a company?
At Russian Appraisal JSC, we’ve been providing valuation and advisory services since 1995. Over three decades, we’ve delivered projects for Rosatom, Gazprom, Lukoil, Russian Railways, VTB Group, Aeroflot, the United Aircraft Corporation, and many other major Russian enterprises. Today, we rank among the top 8 largest valuation firms in Russia. That depth of experience — across business valuation, M&A advisory, financial and tax due diligence, technology audits, and pricing analysis — gives us a grounded perspective on this question.
Over the past two years, we’ve been actively testing AI tools in our practice. In this article, we share what genuinely works and what remains, for now, more marketing promise than reality.
From Spreadsheets to Intelligent Analytics
Five years ago, a valuation professional’s workflow revolved around Excel models and manual data collection. Today, the volume of information required for a single project has grown exponentially: hundreds of M&A transactions, lengthy industry reports, macroeconomic forecasts, competitor profiles, regulatory updates.
This is precisely where AI delivers measurable results. Algorithms can aggregate data from disparate sources, structure it, flag anomalies, and generate preliminary analytics. For a valuation firm, the business case is straightforward: less time spent on preparation and data gathering means more time for analysis and client engagement.
In practice, we already use AI in financial and tax due diligence preparation, historical performance analysis, financial forecasting, and the screening of comparable transactions.
Forecasting: AI’s Greatest Strength — and Its Fundamental Limitation
One of the most promising applications of AI in valuation is forecasting revenue, EBITDA, and cash flows. Models built on time-series analysis, gradient boosting, or neural networks can incorporate historical dynamics and generate scenario-based projections.
In stable industries with rich historical data — retail or telecom, for instance — these tools can be remarkably accurate. The algorithm detects hidden patterns, accounts for seasonality, macro-level influences, and even how a business has historically responded to industry shifts.
The problem, however, is structural: every model learns from the past. When a genuine discontinuity occurs — sanctions, technological disruption, an abrupt change in consumer behavior — algorithms struggle with qualitative context. They don’t understand what they don’t know.
This is why we treat AI-generated forecasts as a starting point. The algorithm produces a range of scenarios; the decision about which scenario is realistic belongs to the expert.
Risk Analysis and Competitive Intelligence
A company’s value is always a function of risk — industry dynamics, regulatory pressure, competitive landscape, macroeconomics all feed directly into the final figure.
AI contributes in two meaningful ways here.
The first is unstructured text analysis. Modern NLP models can monitor news flow, track sentiment shifts, and surface early-warning signals about emerging risks long before they appear in financial statements.
The second is sensitivity modeling. Algorithms can rapidly calculate how a company’s value changes across different scenarios: a rising benchmark rate, currency depreciation, market slowdown. This enhances scenario analysis and helps calibrate discount rates with greater precision.
For the Russian Appraisal JSC team — which regularly works on projects in the oil and gas, financial, and transportation sectors — this means incorporating sector-specific nuance at a level that simply wasn’t feasible before.
The Comparable Company Approach: Better Benchmarking Through AI
The market approach is one of the foundational methods in valuation. But selecting relevant comparable companies and interpreting multiples correctly is time-intensive analytical work.
AI addresses specific pain points here: processing M&A transaction databases, classifying companies by financial and industry parameters, identifying statistical outliers, and mapping correlations between business characteristics and multiple levels.
This reduces subjectivity in peer selection and strengthens the evidentiary basis for conclusions. But the final adjustment of multiples — accounting for the unique attributes of a specific business, strategic factors, management quality — remains the appraiser’s domain.
The Economics of Adoption: What Firms Actually Gain
The primary benefit of AI in valuation practice is speed without sacrificing quality. Automating routine tasks frees experts to focus on what genuinely requires their expertise: strategic analysis, interpreting results, and client-facing work.
This matters particularly in M&A, where timelines are tight. Forbes data suggests that 93% of small business owners believe AI tools help reduce costs and improve profitability. For a consulting firm, the dynamic is similar: more effective resource allocation and a stronger competitive position.
One important clarification: the efficiency gain doesn’t come from replacing the expert — it comes from redistributing their time. Less manual data collection means more capacity for analysis and report quality.
The Limitations No One Mentions in Pitch Decks
AI is not a silver bullet. Three limitations deserve honest attention.
Data quality. An algorithm can process millions of rows of data — but if the underlying inputs are incomplete or flawed, the output will be too. Garbage in, garbage out. Nothing has changed that.
Model opacity. Neural networks can achieve impressive accuracy, but their internal logic is a black box. In valuation practice, where every assumption must be explicitly justified and the reasoning behind calculations must be transparent, this is a serious constraint.
As Alexander Ivanov, Managing Partner of Russian Appraisal JSC, puts it:
“An algorithm only sees numbers. AI can process terabytes of data and surface a million correlations. But it has no concept of market motivation or investment attractiveness. It’s the expert who brings together the data, the context, the experience, and the judgment — and turns all of that into a defensible value conclusion.”
Accountability. Legal and professional responsibility for a valuation rests with the expert — not the algorithm. The appraiser holds the professional judgment, signs the report, and stands behind the conclusions.
Open Source or Proprietary? Choosing the Right Architecture
For any valuation firm, this is a strategic question: adopt ready-made open-source models or build proprietary tools?
Open-source solutions allow you to move quickly and keep initial costs low. But they don’t always reflect the specifics of valuation work, and they can introduce real risks around client data confidentiality. For a firm working with the likes of Rosatom or Gazprom, that’s not an abstract concern.
Proprietary development offers flexibility, algorithmic control, and the ability to align tools with internal standards. But it requires investment, a capable team, and time.
In practice, the optimal path is hybrid: use open-source foundations for standard functionality, and build out the critical components to meet internal standards and regulatory requirements. That’s the approach we’ve taken at Russian Appraisal JSC.
So — Will AI Replace the Appraiser?
The short answer is no.
Business valuation isn’t arithmetic. It requires understanding a company’s strategy, assessing the management team, navigating regulatory risk, valuing intangible assets, and forming a view on long-term trajectory. No algorithm is capable of replacing the professional judgment built through years of practice.
AI is a powerful tool. It accelerates calculations, expands analytical capacity, and reduces costs. But its role is instrumental. The future of the valuation profession isn’t man versus algorithm — it’s the integration of both. Technology and expertise, working together, are what will deliver accuracy, rigor, and reliability in today’s environment.
