AI Is Coming for the Economic Consulting Industry

Artificial intelligence will change the market for economic consultants, likely reducing overall demand and shifting workers to current clients’ in-house units. However, both consulting firms and clients are still studying how to deploy AI, and there may yet be new opportunities for consultants as AI changes the broader economy, write Mona Birjandi and Mery Zadeh.

The economic consulting work behind any major commercial dispute today, whether an antitrust class action, a patent infringement trial, a securities fraud case, or an employment discrimination matter, follows a well-worn pattern: a retained party engages a consulting firm, which assembles a team of economists and quantitative analysts to structure and processes case-relevant data, execute regressions, build damages models, draft expert reports, and presents the results under oath.

This workflow has been largely unchanged for four decades and it has been generously rewarded. The average expert witness fee reached an all-time high of $451 per hour in 2024, up from $200 per hour in 2006, according to ExpertPages.com’s survey of more than 550 experts. SEAK Inc.’s 2024 survey of nearly 1,600 experts similarly found a median of $450 per hour for case preparation and $500 per hour for trial testimony. In major commercial matters, total fees for both sides’ economic experts routinely reach seven figures. A single monopolization case can cost well over $25 million in expert fees alone. Like many industries, though, the advent of artificial intelligence promises to upend the economic consulting model by increasing productivity and reducing the labor required to achieve the same output.

Table 1. Expert Witness Fee Structure Across Practice Areas

One way to see how this model translates into day-to-day practice is through the industry’s standard quality-control protocols. Take, for example, the industry’s “maker-checker“ protocol: every substantive analysis requires an independent economist to replicate the work from scratch before it leaves the firm. This is sound practice, and it effectively doubles the billing cost of every analytical task for senior economists. AI offers a cheaper path to the same standard. Rather than one senior economist auditing another’s work, a swarm of specialized AI agents (e.g., OpenAI’s Agents SDK and Claude Code) can independently check data inputs, audit code logic, stress-test assumptions, and flag inconsistencies with the literature, concurrently, at a fraction of the cost. Realizing those savings will still require the second senior economist to validate the agents’ outputs, identify edge cases and coding errors, or make context-dependent methodological judgments, but this will be a significantly faster process than full replication.

Table 2. How AI will reduce the time to complete different tasks part of a standard consultation

To make these savings in time more concrete, consider a standard task in securities litigation. A securities fraud case requiring an event study (the standard econometric method for measuring price impact) historically required a junior analyst two to four days to collect stock and market data, run Capital Asset Pricing models (CAPM), calculate abnormal returns, and produce tables and exhibits, followed by a separate senior-level audit. Claude Code, Anthropic’s agentic coding tool, can write and execute the entire pipeline in R or Python from a plain-language description, pulling data via API, flagging methodological choices, and producing formatted outputs in under an hour. A second agent can then audit the code for errors, check assumptions against the literature, and run alternative specifications, all before a human economist reviews the output.

How AI changes the demand for consultants

Headcount data from publicly reporting firms confirms that businesses recognize the productivity gains from AI and are replacing workers with it. Charles River Associates (CRA), one of the few publicly traded economic consulting firms, reported a 5.8 percent year-over-year decline in consultant headcount in fiscal year 2024. Meanwhile, revenue per consultant climbed to approximately $727,000, up from approximately $621,000 the prior year, consistent with fewer analysts producing more output per engagement. Both CRA and Cornerstone Research have appointed high-level officers to employ AI.

AI will impact most the market for analysts, who comprise the entry level positions at most consulting firms. Firms that built their model on billing clients for junior hours at steep markups face real pressure. Analysts are mostly responsible for retrieving data, running code, formatting outputs, drafting report sections, the jobs that AI does well and much faster. A tool costing a few hundred dollars a month can replace two weeks of junior work.

The analysts who will remain competitive are those who can supervise AI systems, not merely operate them. The analysts trained in economics and law remain non-negotiable: regression methodology, damages theory, and the ability to evaluate whether a model survives Daubert—the legal standard evaluating the accuracy of expert testimony in court. Those analysts with the training to supervise AI will earn their keep. PwC’s 2025 Global AI Jobs Barometer, analyzing close to one billion job postings across six continents, found that roles requiring AI skills command a 56 percent wage premium over comparable positions.

The question is how these analysts will obtain these skills if consulting firms no longer hire them with the expectation that training will come with the job. Analysts at economic consulting firms are predominantly college and master’s graduates in economics, statistics, or related quantitative fields. Grad school programs at the master level remain the likeliest incubator.

However, if those programs turn out a surplus of analysts with this training, it may depress the wage premium PwC identified. According to the National Center for Education Statistics, U.S. universities award on the order of 30,000 economics bachelor’s degrees annually, with master’s programs in applied economics, data science, and quantitative finance adding further to the pipeline. That supply has not contracted. What has contracted is per-engagement analyst demand at the firms that employ them. More graduates chasing fewer AI supervisory roles means downward wage pressure at entry level that the AI skills premium will only partially offset.

The credentialed-but-not-yet-testifying economist, with five to ten years of experience and who form the middle layer of most consulting firms, will also experience sharply declining demand. Technically solid but not yet scarce, these economists are too expensive to simply manage AI systems but are not yet differentiated enough in expertise to be hired for their advanced skillsets. This tier will compress most sharply as firms field smaller teams per engagement and reduce associate-level hiring even while senior capacity holds.

Indeed, senior experts, the ones who supervise all final case analysis and testify in court, will benefit the most from AI. AI cannot replace expert testimony, and even as per-engagement fee revenue declines as analyst hours thin, the fees for testifying experts may rise as AI-enabled teams perform as well as they do now with less.

The overall equilibrium is a smaller industry by headcount, more concentrated at the senior level, with a hollowed-out middle. Legal practice provides the leading indicator: Harvey AI, now used by over 50 percent of Am Law 100 firms, reports power users saving nearly 37 hours per month. Economic consulting firms are following the same trajectory: Cornerstone Research now deploys an agentic research platform across its case work, applying AI to quantitative analysis, document review, and opposing expert rebuttal, with all outputs replicated by a professional before use in litigation.

Savings may not initially flow to clients, which may eventually further depress demand for consultants

Although AI will increase productivity, how it changes billing rates will be complicated, though the aggregate effect may be to reduce overall costs. On the one hand, if AI cuts analyst hours by 60–70 percent, costs fall and competitive pressure should pass some savings to clients. But the legal profession’s experience is instructive: law firm revenue jumped 13% in 2024 and net income 17% while AI was compressing task time, with efficiency gains accruing to the firms rather than their clients. Every law firm surveyed by BigHand in 2025 acknowledged AI had affected their pricing strategy, but only a third had actually changed their pricing models.

Clients may not yet see lower bills for several reasons. Clients receiving an invoice for hours have no visibility into which hours were AI-assisted, and so firms may still charge clients for ten hours of work when it now takes two with AI. Partners at consulting firms, whose compensation is tied to revenue generation, will be particularly reluctant to reduce fees even as they save with the use of AI.

However, the more durable reasons maintaining current pricing are structural, not informational. First, in high-stakes litigation, clients do not primarily shop on price. They hire the expert with the strongest Daubert track record and the most credible methodology for their specific dispute, which structurally suppresses price sensitivity in ways that do not apply to, say, document review.

Second, market concentration reinforces this dynamic: in practice areas like pharmaceutical patent damages or securities fraud loss causation, the pool of credible testifying experts is small, and firms with established reputations face limited competitive pressure to cut rates regardless of their internal cost structure.

Third, even clients who understand AI’s capabilities may be deliberately cautious about demanding lower fees tied to AI use. If they cannot verify whether AI was used, demanding a discount creates an adversarial dynamic with the expert they depend on in court. Risk-averse clients in high-stakes litigation often prefer to keep that relationship intact rather than press for savings that are difficult to quantify and potentially destabilizing to the engagement.

Some, but not all, of these barriers to lower prices for clients will erode as the clients become more sophisticated. However, it may still be cheaper for many clients to bring these consulting services in house. Already, many corporate legal departments are building internal AI capability specifically to reduce dependency on outside consultants for work that AI can handle. The trend is early but visible. JPMorgan Chase has deployed AI systems to automate legal document review and contract-related work previously performed by law firms and has more broadly moved to replace certain external advisory functions with internally developed AI tools.

Several surveys and industry reports further indicate that reducing outside legal spend is a primary motivation for AI adoption among Fortune 500 legal departments, with in-house teams increasingly building internal capabilities to perform work previously outsourced. The 2025 Chief Legal Officer Survey by the Association of Corporate Counsel found that a majority of large in-house legal departments had begun deploying AI tools for legal research and contract analysis, work previously routed to outside firms on an hourly basis. Darrow AI, a litigation intelligence platform, analyzes regulatory filings and enforcement data to surface viable claims and preliminary damages estimates before a single outside consultant is retained, automating the early-stage case-screening work that law firms historically sent to economic teams. As these in-house capabilities mature, the economic consultant is increasingly engaged only where the analysis must be defensible under oath by a credentialed expert, not merely internally useful. This could end up putting further pressure on incumbent consulting firms though offer job opportunities for displaced analysts.

All is not dire for consultants

In fact, AI simultaneously creates demand for analysts and economists in roles and at stages of the litigation lifecycle that did not previously generate consulting revenue at all.

Perhaps the most overlooked emerging role for economic consultants in the AI era is helping law firms identify which cases to bring and how to prioritize them. Plaintiffs’ litigation finance firms, class action practices, and mass tort teams already use analytics to scan for patterns in public data that indicate viable claims. AI dramatically amplifies this capability. If a new ERISA fee case theory is viable, as illustrated by the 36 percent increase in ERISA class action filings in 2024, an economist can use AI tools to screen thousands of retirement plans for comparable excessive fee structures, rank them by estimated damages, and produce initial case profiles, all before a single retainer is signed.

The trend is early but visible. At firms such as DLA Piper, AI is already being used to analyze large datasets and surface patterns relevant to potential legal violations, reflecting a broader shift toward data-driven identification and development of claims. Platforms like Darrow AI analyze regulatory filings, enforcement actions, and industry data to identify litigation patterns before individual cases materialize. Whether this screening capability migrates permanently in-house or remains the domain of specialized outside consultants, it presents a genuine new revenue stream for analysts willing to work on contingency or with litigation finance backing.

For economic consulting firms, this translates to a new service line: helping corporate legal departments and law firms design and deploy AI-enabled analytics workflows. An experienced damages economist, for example, is well-positioned to specify the parameters of an automated event study pipeline, what data sources to use, which market model is appropriate for which industry, how to flag methodological edge cases, and to certify the output as defensible. The consultant’s role shifts from analyst to architect.   

AI will also present new types of cases that will raise the demand for consultants with expertise in AI. In a 2024 joint statement, the Federal Trade Commission and other antitrust enforcers worldwide signaled increased scrutiny of AI markets and emerging competition risks. , and the Justice Department followed in 2025 with a Statement of Interest arguing that algorithmic price fixing could violate federal antitrust law. These cases require economists who can understand pricing algorithms, simulate competitive counterfactuals, and quantify consumer harm in markets where prices are set by software rather than humans. That is genuinely novel economic work with no established methodological template, which means it commands premium fees and is not easily commoditized.

A separate and growing category of work involves retained economists auditing AI systems for regulatory compliance and disparate impact. This is not litigation work in the traditional sense, it is compliance economics, but it draws on the same methodological toolkit as employment discrimination damages analysis, and it is generating demand from corporate legal and compliance departments that did not exist at scale before AI.

Whether these new lines will offset the compression of traditional billings is still unclear, but they provide some reasons for optimism for consultants who now find their work in a precarious position.

Conclusion

Economic consulting in litigation is not going away. Courts will continue to require credentialed experts to anchor their economic theories. What is changing is the labor content of that work, the structure of the firms that deliver it, and the skill profile of the people those firms need.

It is also too early to know how the market will evolve in response to such a new and flawed technology. AI-generated analyses can hallucinate, fabricating citations, inventing data relationships, or producing plausible but wrong regression outputs. Courts have already sanctioned lawyers for submitting AI-fabricated case law. Uploading case-sensitive data to third-party platforms raises privilege and confidentiality risks many firms are not yet equipped to manage. And if AI eliminates the junior analyst work that traditionally trained the next generation of senior experts, firms may find they have efficient workflows but eventually fewer people qualified to defend them under oath.

But assuming AI can fulfill its potential, industry employment will depends on whether AI-driven cost reductions generate enough new case volume to offset reduced labor demand per case. That demand-side question is an open question. What is not open is the direction of the skill premium: economics and law knowledge retains its value; pure execution without judgment does not.

Authors’ Disclosures: Mona Birjandi is the principal economist and director of data analytics at Outten & Golden, a law firm specializing in labor and employment litigation. Mery Zadeh is the senior vice president of AI governance and risk consulting at Lumenova AI, which consults businesses on AI regulation, compliance, and risk. The authors declare no conflicts of interest. You can read our disclosure policy here.

Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.

Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.