Lawyers embrace litigation analytics but still rely on human judgment

As litigation analytics and artificial intelligence become increasingly embedded in legal practice, some California attorneys say that while the technology improves efficiency and helps inform decisions, it does not replace the experience and judgment needed to advise clients and navigate cases.

That sentiment is reflected in results from Lex Machina's "2026 Impact of Legal Analytics Survey," which the legal analytics platform released last month.

The nationwide survey of 207 law firm professionals found unanimous agreement that litigation analytics add value to their legal practice, up from just over 95% a year earlier.

According to Lex Machina, respondents said they use the data tools to assess case exposure, evaluate judges and opposing counsel, strengthen briefs and motions, and demonstrate expertise to clients, reflecting the technology's growing role in everyday legal work.

However, some California practitioners interviewed by the Daily Journal cautioned that analytics remain only one piece of the decision-making process, particularly as firms increasingly incorporate artificial intelligence into everyday legal work.

"The biggest limitation is litigation analytics are not a crystal ball, and every case is different," said Nathaniel E. Haas, a Los Angeles partner at Watstein Terepka LLP.

While analytics help lawyers make more informed decisions, he noted, "they can never tell you how a judge is going to rule in your case" or eliminate the uncertainty that comes with litigation.

According to the survey results, among firms with more than 50 attorneys, 86% of respondents said they use litigation analytics in practice, while 88% reported clients now expect attorneys to use analytics on their matters.

Adoption remains lower among firms with fewer than 50 attorneys, where 44% reported using litigation analytics, but every respondent in that group likewise agreed the tools add value.

Adam Masarek, a Boston-based attorney and legal data expert with Lex Machina, said those figures continue a trend the company has tracked since first surveying legal practitioners in 2018.

The earliest survey found just 38% of respondents used litigation analytics in their day-to-day work, he said. In recent years, that figure has generally ranged between 60% and 70%.

Masarek said this year's biggest change involved firms seeking to integrate litigation analytics directly into internal systems through application programming interfaces, or APIs, allowing data about judges, venues, opposing counsel and litigation history to work alongside firms' own information and artificial intelligence tools.

He said 73% of respondents expressed interest in that type of integration, a substantial increase from the prior year that he believes is closely tied to AI adoption.

Even as technology evolves, Haas said lawyers must still evaluate whether the underlying data is meaningful before relying on it.

"The usefulness of legal analytics depends on the sample size," he said, explaining that statistics showing how often a judge grants or denies a particular motion may have limited value if they are based on only a handful of cases or include dissimilar matters.

Data quality, he added, is equally important because experienced lawyers have long relied on institutional knowledge about judges and opposing counsel, with modern analytics simply placing more empirical support behind practices litigators have followed for decades.

Haas said his firm uses litigation analytics throughout the life of a case, from motions to dismiss and class certification through summary judgment and appeals, helping clients evaluate litigation risk and settlement exposure. The information can provide reassurance that pursuing a particular motion is worthwhile, but it never dictates strategy by itself.

"Good lawyers manage that uncertainty," Haas said. "You never promise you're going to win a motion because judges are human and they might disagree with you or they might make a mistake."

Masarek said those practical uses largely mirror what firms have reported to Lex Machina over the years.

Claim assessment and case strategy remain among the most common uses for litigation analytics, he said, helping attorneys decide where to file suit, whether to remove a case, pursue motions or consider settlement.

Firms also increasingly rely on analytics for business development, using objective litigation data to demonstrate experience in client pitches, respond to requests for proposals, identify prospective clients and even evaluate potential lateral hires.

The survey likewise found firms increasingly are using analytics before litigation even begins. Across firms of all sizes, respondents identified pitching clients and demonstrating expertise as the leading business-development use for litigation analytics, while firms also reported using the data to identify opportunities and better understand litigation trends.

Criminal defense attorney Dmitry Gorin, an Eisner Gorin LLP partner, said artificial intelligence has already transformed legal research by allowing lawyers to summarize lengthy records, organize discovery, compare testimony and identify relevant authority in minutes rather than hours. However, he said speed should not be confused with accuracy.

"AI increases efficiency, but doesn't replace judgment," Gorin said, adding that every citation, factual assertion and legal conclusion still must be verified by an attorney.

In criminal defense, Gorin said technology can organize information, but it cannot evaluate witness credibility, predict juror reactions or gauge the effectiveness of cross-examination.

"Courtrooms involve human beings, not algorithms," he said.

"The danger of AI is that bad data produces bad conclusions. If the underlying information is incomplete or inaccurate, the resulting analytics become misleading. Lawyers should avoid treating statistics as predictions."

He also cautioned attorneys to ensure AI platforms comply with ethical obligations governing privileged client information.

Masarek likewise said the next major evolution for litigation analytics will be determining how empirical legal data is incorporated into AI systems. As large language models become more common in legal practice, he said, attorneys increasingly will expect those tools to draw on reliable litigation data rather than simply generate text.

For Haas, the profession's future remains grounded in the same principle that has guided lawyers for generations.

"As long as humans are still the judges deciding issues, and as long as humans still sit on the juries that decide cases, the legal analytics tool will probably always get better, but they will never replace what lawyers have done for hundreds of years," Haas said.