Not long ago, one of us (Michael) had lunch with a friend who has spent his career as a plaintiff’s lawyer. Michael asked how his practice was going and was told his caseload—much of it plaintiff-side employment law—had doubled. And yet, despite that doubling, he had more free time than he’d had in years.

When Michael asked how both could be true, his answer was simple: AI.

That answer captures a shift Minnesota’s business community should understand: AI technologies are increasingly taking sides in litigation.

We build AI ourselves. Cloud Court, the Minneapolis-based AI company we co-founded nearly a decade ago, develops technology focused on testimonial evidence. Some of our technologies simply make time-consuming work faster. Other capabilities are more unusual: predicting what an adverse attorney or witness is likely to do or say before they say or do it. We work closely with corporate legal departments and law firms of all sizes, representing plaintiffs and defendants alike.

We build tools; we don’t take sides. We’re essentially Switzerland, and that neutrality gives us a useful vantage point on a legal tech market that is becoming increasingly partisan.

Neutral tools and partisan tools

For most of history, the legal profession has not distinguished itself by embracing new technology. When it has, the most consequential tools have generally been party-neutral. Legal research moved from law libraries to online databases; document review went digital, replacing manual review of oceans of paper. These technologies are non-partisan. Plaintiffs and defendants use them alike. But the nonpartisan nature of emerging technologies has been changing.

Since the emergence of large language models, there has been an explosion of new tools and technologies entering the market. Among the many entities that attempt to track this, Stanford’s CodeX TechIndex now lists more than 5,000 legal-technology companies. Another entity, the Legaltech Hub, catalogs roughly 3,000 solutions. By March 2026, more than 1,000 of those solutions incorporated generative AI.

A new category is emerging alongside those predominantly neutral technologies: AI tools built specifically for one side of the “v.” Among them, considerably more are built for plaintiffs than for defendants.

Follow the money

Plaintiff-focused technologies have been attracting big investments. In October 2025, EvenUp, which builds AI for personal-injury law firms, raised $150 million in a Series E at a valuation exceeding $2 billion. A week earlier, Eve, an AI platform expressly built for plaintiff-side law firms, announced a $103 million round at a valuation above $1 billion. Thus, within days of each other, two plaintiff-focused legal-AI companies based in the Bay Area had attracted more than a quarter-billion dollars.

They’re not alone. Supio, which describes its platform as built exclusively for plaintiff law, has raised $91 million. Darrow, which uses AI to detect legal violations and route the resulting plaintiffs to firms, has raised around $80 million and says it has connected more than 100,000 plaintiffs to cases. Behind them sits a pipeline of younger companies: AlphaLit, Andco, Finch, Kalinda. And case-management vendors are building AI into software that plaintiff firms already run.

What matters is not just how much money is flowing but what that money is buying. Plaintiff-oriented tools attack the bottlenecks that historically limited caseloads: intake, claimant screening, medical-record review, chronologies, demand packages, discovery, and case management. Reduce the labor required to evaluate and prosecute a claim, and you do more than just make existing cases cheaper. You make previously uneconomic cases worth taking. AlphaLit, which raised $3.2 million in January, states the premise outright: Most prospective plaintiffs’ calls go unanswered because vetting a small case costs more than the case is worth.

That dynamic is particularly visible in employment and personal injury litigation, where lawyers must sift through emails, texts, personnel records, policies, medical records, and competing narratives before deciding whether a claim is viable. AI dramatically reduces the cost of that preliminary work and, with it, the economic threshold for accepting a case.

Why the tilt?

To be clear, AI is being built for the defense side, too, but the tools and incentives look different.

Defense-oriented technology more often focuses on controlling litigation costs through bill review, matter management, counsel selection, and case valuation. Some of it is extremely helpful. But three factors push the “partisan AI” market toward plaintiffs.

The first factor is compensation. If technology can turn 10 hours of work into two, plaintiffs’ firms win for the simple reason that technology that lets them handle more cases directly increases revenue. Contingency fee engagements incentivize efficiency and the tech adoption that delivers it. The same is not true of defense firms, which still often (though not always) bill based on time spent. For defense firms compensated in six-minute increments, it makes little sense to pay for technology that results in eight fewer hours billed, at least in a universe where clients fail to recognize and reward that efficiency.

The second factor is adoption. A small plaintiffs’ firm may have no IT department and no tech committee, with clients who are utterly indifferent to the tools their lawyers use. If the partners decide a product will help them take more cases and resolve them faster with larger recoveries, “due diligence” and adoption can take an afternoon.

Large defense firms operate in a different world. To limit reputational risk (stories of firms’ use of AI-hallucinated legal “authorities” are legion), new tools must clear a byzantine gauntlet of technology, privacy, security, procurement, and ethics reviews. They must also satisfy the AI policies of multiple large corporate clients, which (while well meaning) often conflict. As a result, deployment takes months, often longer, and by then newer tools have superseded the old. To navigate that, many firms now standardize on a few broad platforms, such as Harvey and Legora, rather than niche tools with deeper capabilities. Plaintiffs’ firms face none of this, so they move faster.

The third factor is AI developer and investor economics. Legal-AI startups live on short runways. They need customers who can buy, deploy, give feedback, and expand quickly. Small, entrepreneurial firms (many plaintiff firms) fit that profile. Enterprise defense procurement, operating at a more glacial pace, generally doesn’t. Tech companies and their investors respond to incentives just as lawyers and companies do, which is why they are investing heavily in technologies directed to plaintiffs.

What companies can do

Nearly every law firm in the litigation system, plaintiff or defense, does better when more lawsuits are filed. There is one notable exception: the companies being sued. For them, litigation remains an unwanted cost of doing business, and most signs point toward more of it, not less.

Corporate clients are not powerless. But if they want to capture the efficiencies and advantages of emerging technology, they will have to change how they do business with their outside counsel.

Ford Motor Company offers an assertive model. Darth Vaughn, who leads legal innovation and operations at Ford, tells outside firms that gone are the days when they merely competed with one another. Now they also compete with Ford’s own legal team, which is building and deploying AI tools to bring as much work in-house as possible. Vaughn’s challenge to outside counsel: What can your law firm do that we cannot now do ourselves? That question is getting harder to answer.

For Minnesota companies wondering where to start, a sensible place is their own billing records. One of our clients analyzed $250 million of its litigation spending and found, to its surprise, that at least 26% of that spend went to testimony-related tasks—many of them recurring, labor-intensive, and well suited to AI. That kind of analysis turns a vague efficiency problem into concrete opportunities. Which tasks, repeated across matters and outside firms, consume the most attorney time? The answers will vary by company, but scrutinizing your own data is a solid starting point before surveying the growing blizzard of new AI tools.

Efficiency, however, is only part of the equation. Some litigation technologies matter not because they turn 10 billable hours into two but because they help lawyers better understand the evidentiary record, identify strengths and weaknesses earlier, and position a case more effectively for resolution. Those tools can be harder to measure in saved hours, but their economic impact (cases resolved or won) may be far greater.

Back to lunch

Michael’s friend’s story does not prove a national trend. But it is a useful picture of the economics now taking shape.

Plaintiffs’ firms are presented with technology that helps them screen more claims, take more cases, and prosecute those cases with less human labor. Corporate defendants operate through a different ecosystem, in which efficiency can conflict with hourly billing and new technology can take months to approve.

That asymmetry may not last. But companies should not assume their outside law firms will eliminate it on their own. Corporate legal departments should experiment with emerging technologies themselves, identify the tools that actually create efficiencies and advantages, and then require their outside counsel to use them.

The question is no longer whether AI will change the economics of litigation. It already is. The question for companies is whether they will capture those efficiencies as quickly as the lawyers bringing cases against them.