The transformation of legal technology

For most of its history the legal industry bought technology at the edges: billing, document storage, e-discovery. That has changed. Generative AI has moved into the centre of legal work itself — research, drafting, review, and increasingly the intake and triage that decides what reaches a lawyer at all.

The spending follows. Law firm technology investment grew 9.7% in 2025, and knowledge-management spend grew 10.5%, in what the Thomson Reuters Institute and Georgetown Law describe as one of the sharpest sustained increases the sector has recorded.

What makes this different from previous cycles is who is buying. Corporate legal departments more than doubled their generative AI adoption in a single year, from 23% to 52%, and 64% now expect to rely less on outside counsel as a direct result.

Legal technology took $4.08bn of venture funding in 2025 on Crunchbase's count — a 77% increase on the year before, and a record. Specialist trackers count differently: Legaltech Hub logged $4.28bn across 107 rounds, while Artificial Lawyer's wider definition reached $5.99bn. All three agree on the direction.

The shape of that money matters more than the total. Artificial Lawyer counted fourteen rounds of $100m or more in 2025, while the number of companies funded fell 27% and the number of active investors fell 31%. Capital is consolidating behind a small group of category leaders.

Valuations have followed. Harvey raised $200m at an $11bn valuation in March 2026, having been valued at $8bn only months earlier. Legora reached $5.5bn in April 2026, up from $1.8bn the previous October. Clio's $1bn acquisition of vLex was the largest private legal-tech deal on record.

The counterweight is the exit market: total legal tech exits were worth $2.29bn in 2025, down 39% year on year. Money is going in faster than it is coming out, which is exactly the condition that funds aggressive hiring.

Thomson Reuters' 2026 Future of Professionals survey found 74% of legal and professional-services respondents using AI several times a week, and 44% using it multiple times a day. In the UK, LexisNexis recorded generative AI use rising from 46% to 61% of lawyers in nine months, with the share of firms reporting no AI plans at all collapsing from 15% to 6%.

Firm size still decides the pace. The ABA's technology survey found US law firm generative AI adoption tripling from 11% to 30% year on year — but 46% at firms of 100 or more lawyers against 18% among solo practitioners.

Strategy matters more than tooling. Organisations with a formal, visible AI strategy are 3.9 times more likely to realise the benefits they set out to achieve, and firms with broad adoption are nearly three times more likely to report revenue growth.

The governance gap is equally real: 34% of professionals report using AI tools their organisation cannot monitor, rising to 41% where staff feel the firm is moving too slowly. That shadow usage is one of the clearest drivers of demand for AI governance and legal engineering hires.

Bloomberg Law reported in 2026 that at least sixteen leading firms had twenty-five or more AI-related roles open simultaneously, and that these have become among the hardest positions in the profession to fill. Firms are asking for a decade of applied AI experience alongside a record of driving adoption inside a professional-services business — a combination that barely exists.

Compensation reflects the scarcity. Published benchmarks include Pillsbury at up to $440,000 for a director of data science and AI engineering, Covington & Burling at $438,000 for a director of applied AI, and Latham & Watkins at $295,000 to $400,000 for AI governance and innovation leadership.

The Chief AI Officer has become a real title in law. Littler appointed one in April 2026; Pillsbury created the role for Oz Benamram; Ropes & Gray brought in a chief of AI strategy from Meta; Husch Blackwell hired its senior director of AI advisory out of Adobe.

Talent expectations have flipped too. Almost a quarter of legal professionals say they would turn down an offer from a firm without professional-grade AI tools, and 24% of those experiencing an AI capability gap are considering leaving within two years — at an estimated replacement cost of $232,000 per professional.

Gartner expects roughly half of all contract reviews to be handled by self-service systems by 2029, escalating only one in ten to a human, and 60% of legal departments to run AI-driven intake that resolves half of all requests without lawyer involvement. In-house legal technology budgets are forecast to double by 2028.

It will not be a smooth line. Gartner also expects more than 40% of agentic AI projects to be cancelled by the end of 2027 on cost or unclear value, and Forrester anticipates enterprises deferring a quarter of planned AI spend into 2027 over return-on-investment concerns. Both are arguments for hiring people who can tell the difference between a demo and a system.

Regulation is the other engine. The EU AI Act's high-risk obligations were pushed to December 2027 under the May 2026 Digital Omnibus, with embedded-product AI following in August 2028. Penalties reach €35m or 7% of global turnover. That work lands on legal operations, and it needs people who understand both the statute and the stack.

Accuracy remains the open question. Independent benchmarking put Harvey at 94.8% on document question-answering, with AI six to eighty times faster than lawyers across tested tasks — while Stanford research found error rates of 17% and 34% in two major legal research assistants. Verification is becoming a discipline of its own.