The year 2025 marked a genuine inflection point in legal technology – the moment when AI moved from experimental novelty to operational necessity, when a billion-dollar deal reshaped the competitive landscape, and when regulatory reform appeared to gain renewed momentum.

As I look back over the stories I have covered this year, they chronicle an industry transforming at unprecedented speed, with agentic AI emerging as the defining concept, record-breaking consolidation redrawing market boundaries, and access to justice advocates proving that AI adoption could be both rapid and meaningful.

What follows are the 10 trends that defined legal technology in 2025, based primarily on my own blog posts, podcast interviews and event coverage throughout the year. These are not ranked – each represents a significant trend that reflects how technology is reshaping the legal profession.

1. Agentic AI becomes the industry’s north star.

The dominant theme of 2025 was the emergence of agentic AI – autonomous systems capable of multi-step reasoning, self-evaluation and complex workflow execution, all without constant human prompting. Agentic AI marked a significant step forward from the chatbot-style AI that characterized 2023 and 2024.

This kicked off in January, when LexisNexis released its AI-powered legal assistant Protégé to general availability, offering both generative and agentic AI capabilities, including autonomous task completion and the ability to review its own work. LexisNexis expanded on this in March, introducing what it said was the first personalized voice AI assistant for legal work, along with a set of advanced reasoning models that included a Planner Agent to break down complex legal questions, an Interactive Agent to modify the agent’s plan, and a Self-Reflection Agent for evaluation and refinement of the agent’s work. By August, LexisNexis added Protégé General AI, offering access to multiple general-purpose AI models alongside legal-specific AI, with agentic workflows including AI-guided research and an “Orchestrator Agent.”

In June, Thomson Reuters unveiled agentic CoCounsel, describing it as a “fundamental shift from AI assistants to intelligent agentic systems,” with guided workflows for document drafting, policy generation and deposition analysis, and teasing the release of its next-gen agentic CoCounsel for legal. In August, it delivered on that promise, launching CoCounsel Legal, a new platform that combined agentic workflows with deep research capabilities, all grounded in Westlaw content, which the company positioned as the most comprehensive artificial intelligence solution for legal professionals to date.

At ILTACON in August, I reported that agentic AI was the “buzzword of 2025,” appearing in virtually every major product announcement. Litera unveiled Lito, its agentic assistant. NetDocuments introduced AI-powered document profiling and an agentic AI tool for editing documents in Word. The shift was captured by LexisNexis’s stated goal: Protégé would automate 15-20% of lawyer tasks by 2028 – providing a concrete metric for agentic ambition.

The year ended with Harvey’s December Reddit AMA, where the company, whose valuation grew this year to $8 billion, described its 81% increase in daily active usage since 2023 and outlined plans for even deeper agentic capabilities.

Clearly, by the end of 2025, the question was no longer whether AI could help lawyers, but rather how much of a lawyer’s work it could complete independently.

2. Clio and vLex create a new category with a billion-dollar deal.

On June 30, Clio announced its $1 billion acquisition of vLex – the largest M&A transaction in legal tech history for a privately held company. The deal, which closed in November following Spanish regulatory approval, combined Clio’s practice management dominance among smaller firms – and increasingly, larger ones as well – with vLex’s AI-powered global legal research platform and its existing deployment at eight of the world’s 10 largest law firms.

When the deal was first announced, Clio founder and CEO Jack Newton characterized it as a “watershed moment” that created “a new category of intelligent legal technology at the intersection of the business and practice of law.” The combined company commanded $400 million in annual recurring revenue, served 400,000 legal professionals, and achieved a $5 billion valuation through a simultaneous $500 million Series G funding round (with an additional $350 million in debt financing from Blackstone and Blue Owl).

At his ClioCon keynote in October, Newton unveiled Clio’s “Intelligent Legal Work Platform” vision – AI that would “run overnight” to extract deadlines, update calendars and manage tasks proactively. The new product lineup included Clio Work, Vincent by Clio, and Clio for Enterprise. As I reported after the keynote, the audience was “shell-shocked” by the scope of announcements.

The significance extended beyond this one company. The acquisition, Newton noted, made Clio “one of only three companies in the world, alongside Thomson Reuters and LexisNexis, with such depth of legal data.” What had been a clear division – Thomson Reuters and LexisNexis for research, Clio for practice management – collapsed into a three-way competition for the integrated legal platform market.

3. The Harvey-LexisNexis alliance signals a new competitive era.

In June, Harvey and LexisNexis announced a strategic partnership that industry analyst Richard Tromans called “possibly the most important legal tech move in a decade.” The alliance integrated Harvey’s AI platform with LexisNexis’s Protégé, primary law and Shepard’s Citations, and set the stage for co-developing workflows for motions to dismiss and summary judgment.

The deal came soon after Harvey raised $300 million in Series E funding at a $5 billion valuation, with RELX (LexisNexis’s parent company) participating as an investor. In a July podcast interview, LexisNexis CEO Sean Fitzpatrick explained the partnership was driven by customer demand from large law firms seeking Harvey’s capabilities within LexisNexis’s trusted content environment.

LawNext guest columnist Ken Crutchfield offered a provocative analysis in September: Harvey was “playing a different game,” likening its strategy to Oracle’s 1980s database dominance. Harvey’s competitive advantages, Crutchfield argued, extended beyond product features to reference accounts (Allen & Overy Shearman, Paul Weiss, PwC), powerhouse investors, and a top-down partner adoption approach that created first-mover advantage with senior lawyers.

By December, Harvey had raised another $160 million, bringing its total raise — just during 2025 — to $760 million, and its valuation to a whopping $8 billion, and the company’s two founders held a Reddit AMA defending its growth. The competitive implications were clear: Thomson Reuters had its own AI strategy through CoCounsel; Clio had acquired vLex; and Harvey-LexisNexis had formed an alliance. The market, it seemed, was consolidating around a handful of major platforms.

4. AI adoption rates double across the profession.

If 2024 was the year of AI experimentation, 2025 was the year of deployment. Multiple surveys documented adoption rates that doubled or tripled year-over-year:

  • Thomson Reuters’s April 2025 survey found 26% of legal organizations actively using generative AI, up from 14% in 2024. The mood shifted dramatically: 55% expressed excitement or hopefulness compared to 35% who reported hesitancy in 2024.
  • The ABA’s Legal Technology Survey Report, released in March, showed AI adoption tripled from 11% to 30% in one year, with large firms at 46% and solo practitioners at 18% (up from 0% in 2022).
  • Smokeball’s 2025 report, published in March, found generative AI adoption among small firms nearly doubled to 53% from 27% in 2023.
  • The Washington State Bar Association’s survey, released in June, found that only 25% of the state’s lawyers used generative AI applications in their practices on a regular basis.
  • Everlaw’s e-discovery survey, released in July, showed 37% of professionals using AI—up from 12% two years prior. The report noted that generative AI achieved in three years the adoption level that cloud computing took a decade to reach.

Yet the surveys also revealed an implementation gap. Thomson Reuters found only 41% of organizations had generative AI policies and only 40% provided training. Just 20% were measuring ROI. A survey report by AllRize identified a disconnect: 89% of firms relied on Microsoft productivity tools, but most had not connected AI with existing infrastructure.

(To see how these surveys compared against each other, see my story, An AI-Assisted Look At Four New Surveys On AI Adoption In Law: How Do They Compare? Differ?)

The trajectory was unmistakable – AI was becoming standard practice – but strategic adoption remained a differentiator.

5. Legal aid leads AI adoption, proving access-to-justice potential.

Perhaps the year’s most surprising development was that legal aid organizations adopted AI at nearly twice the rate of the broader profession. A survey conducted in May by Everlaw, NLADA, Paladin, and LawSites found 74% of legal aid organizations using AI versus 37% of the general legal profession. Daily or weekly usage was reported by 66% of respondents.

The motivation was clear: 88% believed AI could help address the justice gap, and 90% said AI’s full potential would enable them to serve more clients. Thomson Reuters’s “AI for Justice” program, which provided CoCounsel to legal aid organizations, released one-year results in October showing attorneys saving up to 15 hours per week, organizations serving 50% more clients daily, and urgent case materials prepared 75% faster. The Legal Aid Society of San Bernardino – 45 staff serving more than 9,000 people annually—exemplified the impact.

New tools emerged specifically for underserved populations. LawDroid launched LawAnswers AI in September, a nationwide platform addressing the reality that 92% of low-income Americans cannot get legal help. The American Arbitration Association announced an AI-powered chatbot for self-represented parties, providing plain-language guidance for non-lawyers navigating arbitration.

Rebecca Sandefur and Matthew Burnett, interviewed on the LawNext podcast in September, framed the issue starkly, noting that 120 million legal problems go unresolved annually in America. Their “Justice Workers” initiative argued that the access-to-justice crisis was actually “a crisis of democracy.”

6. Regulatory reform sees renewed momentum.

Washington State became the third jurisdiction to fundamentally reform legal services regulation when, in the closing days of 2024, the Washington Supreme Court signed an order establishing a 10-year pilot allowing non-lawyer-owned entities to practice law.

Following Utah and Arizona, Washington joined a growing movement to reconsider Rule 5.4’s prohibition on fee-sharing with non-lawyers.

A June Stanford study provided the most comprehensive assessment yet of these reforms. Arizona’s program showed significant growth: from 19 entities in 2022 to 136 entities by April 2025 — a six-fold increase. Notably, 85% of Arizona entities served individual consumers, not corporations, and 134 of 136 employed lawyers for legal services. Consumer harm was minimal: Utah reported only 20 complaints across all entities, a harm-to-service ratio of 1:5,869.

Utah also moved on bar admission, announcing in October a new skills-based pathway (Rule 14-703A) which begins accepting applications this month (January 2026). The alternative requires 240 hours of supervised practice, skills coursework, and a written assessment—addressing financial barriers to traditional bar admission.

The broader context: four states now offer skills-based admission, with five more running pilots.

The Association of Professional Responsibility Lawyers called on the ABA to reform Model Rule 5.4, citing international models and domestic evidence that “no doomsday scenario” had materialized. Washington State Bar Association Executive Director Terra Nevitt captured the moment: “A lot has changed in 100 years, but not when it comes to legal regulation. Now we are at a regulatory crossroads.”

7. Conferences signal an industry at an inflection point.

The 2025 conference circuit reflected an industry in transformation. At ClioCon in Boston in October, 2,700 attendees witnessed what felt like a historic moment, as CEO Jack Newton’s keynote unveiled the company’s vision for the Intelligent Legal Work Platform. Futurist Richard Susskind delivered the closing keynote, reinforcing the theme of fundamental change.

In August at National Harbor, ILTACON featured 233 vendor booths and 27 startups, with me reporting the overall sense of the conference as “a new dawn in legal tech.” During or around ILTACON, Thomson Reuters launched CoCounsel Legal with agentic capabilities; LexisNexis unveiled Protégé General AI; and product demos across the floor were “paradigm-shifting.”

At Relativity Fest in Chicago in October, the conference built on the company’s announcement earlier in the year of a 2028 cloud mandate – all new matters must be on RelativityOne starting Jan. 1, 2028 – marking a definitive end to the on-premises era. The company also moved aiR for Review and aiR for Privilege into core subscriptions at no additional cost.

At the Everlaw Summit in October in San Francisco, the company revealed a similar democratization strategy: three AI features previously sold separately became included in base subscriptions. After an eight-month beta with more than 25 million documents reviewed, Deep Dive achieved general availability.

The LSC Innovations in Technology Conference last January provided counterpoint, highlighting the funding gap: while legal tech raised approximately $4.98 billion in 2024, it was mostly for large-firm and corporate tools. The contrast between Big Law innovation and legal aid reality remained stark.

8. Legal tech funding continues record pace.

Investment dollars continued flowing into legal technology at remarkable rates. Beyond Clio’s $500 million Series G, the year’s major funding rounds included:

VC firm The LegalTech Fund closed its second fund at $110 million – nearly four times its inaugural $28.5 million 2021 fund. Strategic investors included Clio, DocuSign, Harbor, Orrick, and Thomson Reuters Ventures. I noted the close was “particularly notable given the challenging venture capital environment.”

Private equity remained active. BayPine Capital took a majority stake in Harbor in June, helping to fuel nn acquisition spree. ProfitSolv secured investment co-led by FTV Capital and Lightyear Capital. Oakley Capital, exiting its vLex position, chose to roll significant proceeds into Clio equity rather than taking cash — signaling institutional confidence in legal tech’s trajectory.

9. AI benchmarking establishes credibility through measurement.

The Vals Legal AI Report (VLAIR) benchmark study, released in February, represented the first independent, systematic effort to evaluate legal AI tools against human lawyers. Developed in partnership with Legaltech Hub and a consortium of law firms that included Reed Smith, Fisher Phillips, McDermott Will & Emery, and Ogletree Deakins, it tested Harvey, Thomson Reuters CoCounsel, vLex Vincent AI, and Vecflow Oliver across six task categories.

Key findings established notable baselines: Harvey achieved the highest scores in five of six tasks, including 94.8% accuracy for document Q&A. CoCounsel achieved the highest average score (79.5%) across four tasks. AI tools operated 6-80 times faster than lawyers. LexisNexis withdrew from most tasks after report completion.

In October, a second Vals AI study, this time of legal research tools, suggested that both legal-specific and general large language models are now capable of performing legal research tasks with a level of accuracy equaling or exceeding that of human lawyers.

The Bluebook’s 22nd edition, released in September, provided the first standardized citation format for AI in legal research (Rule 18.3). Yet the rule sparked controversy: critics argued it lacked guidance on when AI citation was appropriate, raising questions about attorney technological competency.

At ILTACON, LexisNexis articulated six principles for “courtroom-grade AI”: grounding in authoritative content, verifiable citations, continuous updates, transparency, explainability, and enterprise security. The framing reflected a maturing industry recognizing that trust required measurable standards.

10. Platform convergence reshapes the competitive landscape.

The year’s overarching narrative was platform convergence – the collapse of traditional boundaries between practice management, legal research, document management and AI capabilities. Clio’s Jack Newton articulated the shift explicitly: the division between “business of law” and “practice of law” software was becoming obsolete.

This was not mere marketing hype. Clio’s acquisition of ShareDo in March (later rebranded as Clio Operate) brought enterprise case management to a company long known for a smaller-firm focus. NetDocuments launched ndConnect in July to integrate third-party AI (initially Harvey and Legora) within its document management security framework. SimpleDocs acquired Law Insider in September, combining document automation with the world’s largest contracts database.

Thomson Reuters’s agentic CoCounsel worked across Westlaw, Practical Law and enterprise content. LexisNexis’s December announcement of Protégé Next Generation unified legal content, customer documents, and web insights with automatic “Best Fit” model selection. The American Arbitration Association launched an AI arbitrator for construction disputes – AI evaluating case merits and generating recommendations, with humans validating final decisions.

The emerging competitive structure is characterized by integrated AI/legal research platforms such as Thomson Reuters, LexisNexis and now Clio, Harvey as a specialized AI provider with strategic partnerships, and specialty vendors facing increasing pressure to integrate or be acquired.

Not coincidentally, the year ended with Fastcase filing a lawsuit against Alexi over data licensing – a reminder that consolidation creates not just opportunity, but also friction.

Bottom Line

What distinguished 2025 was not just the pace of change but its coherence. The year’s developments offered a clear vision of what lies ahead: AI agents handling substantial portions of legal work; a handful of major platforms competing for comprehensive market share; regulatory frameworks adapting to enable innovation; and access-to-justice applications proving AI’s potential for societal benefit.

The questions for 2026 are whether these trends accelerate or encounter friction. Will agentic AI deliver on its promise of autonomous task completion? Can regulatory reform spread beyond the current handful of states? Will the legal aid sector’s AI leadership translate into measurable justice gap reduction?

The industry I have chronicled for three decades stands at what I and others believe is an inflection point. The technology exists. The investment exists. The regulatory openings exist. What remains is execution – and the profession’s willingness to embrace fundamental change.

Photo of Bob Ambrogi Bob Ambrogi

Bob is a lawyer, veteran legal journalist, and award-winning blogger and podcaster. In 2011, he was named to the inaugural Fastcase 50, honoring “the law’s smartest, most courageous innovators, techies, visionaries and leaders.” Earlier in his career, he was editor-in-chief of several legal publications, including The National Law Journal, and editorial director of ALM’s Litigation Services Division.