Highlights
- Clients are adopting AI faster than law firms, creating a widening gap in expectations and preparedness.
- The firms seeing real traction lead with culture and curiosity, not just buying the latest technology.
- Structural gaps remain — from the SQE's silence on AI to slow-moving regulation — even as adoption accelerates.
The question used to be whether AI belonged in law. That debate is over.
At the Thomson Reuters British Legal Tech Forum Insight Stage, a full day of conversations with law firm leaders, in-house counsel, legal educators, and solicitors made one thing abundantly clear: the industry has moved on.
The new, more challenging question is whether firms are building the right foundation before the window closes.
Jump to ↓
The three-stage AI adoption model
The data doesn’t flatter the industry
What practitioners are actually experiencing
The people problem is the real problem
The strategic window is open — for now
The three-stage AI adoption model
A useful frame is a three-stage model: experimentation, systemisation, and transformation.
Most firms spent 2024 and 2025 in stage one — tinkering, testing, exploring. The majority are now somewhere in stage two, trying to move from “let’s try this” to “this is how we work.” Stage three, where AI genuinely reshapes a firm’s business model, is visible on the horizon — but not yet reached by most.
The firms that get there won’t be the ones who bought the most technology. They’ll be the ones who made deliberate choices whilst others were still deciding. We’ve written about this shift in detail — and the competitive window is narrowing faster than many firms expect.
The data doesn’t flatter the industry
The Thomson Reuters Institute’s 2026 Generative and Agentic AI in Professional Services report offers an honest picture of where things stand. The headline that should make every managing partner uncomfortable is this: clients are now moving faster towards AI than law firms are.
A few other findings from the report are equally striking:
- AI adoption has nearly doubled year-on-year in terms of central workflow integration
- Fewer than one in five firms are measuring ROI — and those that do focus on operational metrics, not business outcomes
- Legal research is now the number one generative AI use case in law firms used by 80% of those deploying the technology — overtaking document review and summarisation for the first time
- CoCounsel, Thomson Reuters’ AI platform, has surpassed one million users globally
Adoption is real. But most firms don’t know whether it’s working, and most clients don’t even know whether AI is being used on their matters. That communication gap creates both reputational risk and a missed opportunity to demonstrate value.
What practitioners are actually experiencing
The sharpest insights at British Legal Tech Forum came not from the data but from practitioners willing to speak plainly about what’s working and what isn’t.
Adam Kerr, Managing Partner at Primus Law, described his firm’s early AI journey with characteristic candour. His team had expected something out of a science-fiction film, and the early tool — whilst genuinely useful — wasn’t quite that. “We probably hadn’t got the messaging quite right,” he admitted. The lesson wasn’t that the technology was wrong. It was that expectation management matters as much as implementation.
The firms seeing real traction shared a common trait: they led with culture before tools. The analogy that resonated most was treating AI like a brilliant but junior colleague — give it good context and precise questions, and you get excellent output. Give it vague prompts and expect magic, and you’ll be disappointed.
The winners in this industry won’t be the people who buy the latest tech. It’s the people who buy the best tech to solve their problems.
Director of Product Solutions, Thomson Reuters
The in-house perspective added another dimension. Accenture’s legal team — 3,300 professionals across 42 countries — deployed CoCounsel to its entire function and is already seeing a shift in how external counsel is instructed. Research that previously went to outside lawyers as broad questions is now done internally first, with instructions becoming sharper and more targeted. One senior lawyer in Germany cut a complex litigation response from a full day to half that time.
The philosophy Accenture is building towards is “human in the lead” — not a human checking every AI output, but a human who has designed the system, set the guardrails, and is accountable for what it produces.
On pricing, the conversation was refreshingly honest. The legal profession has spent decades billing by the hour. Now that AI compresses time, clients are asking reasonable questions about whether costs should follow. Some firms are already moving towards output-based models.
The seven strategic rules for legal AI success offer a practical framework for firms navigating exactly this tension — how to capture ROI without eroding the trust that underpins the client relationship. Those who get ahead of it will be better placed than those who wait for clients to force the issue.
The people problem is the real problem
Across every session at British Legal Tech Forum, one theme dominated: the technology is the easy part. The harder work is the human change that surrounds it.
Defining the AI-ready lawyer turned out to be less about technical fluency and more about a single trait: curiosity. “An AI-savvy lawyer in one word to me is curious,” said one legal educator. The willingness to explore, try something that might not work, and ask better questions — that’s the foundation everything else is built on.
That curiosity is unevenly distributed, and candidly, unevenly expected across seniority levels. Junior solicitors are often the most enthusiastic adopters, but their energy isn’t always reciprocated by the senior lawyers who set the culture.
One firm has addressed this directly by creating an advisory board of trainees and apprentices who actively shape the firm’s AI learning strategy — flipping the traditional direction of knowledge transfer.
There are structural gaps worth naming too:
- The SQE still contains no reference to AI. The solicitors’ qualifying examination — the profession’s entry standard — hasn’t caught up with the reality of practice
- The regulatory framework is starting to move, but slowly. The Ayinde judgment, the Civil Justice Council’s recent consultation on AI disclosure requirements in litigation, and growing SRA scrutiny of hallucinated case citations are all signals of what’s coming. CoCounsel’s citation-checking tools are specifically designed to address this risk — and given there are now over 50 logged instances of hallucinated cases in UK courts, the need has never been more pressing
- Some firms are already allocating 20% of associate time specifically for AI exploration — a clear signal of where career development is heading. The rise of embedded AI that makes adoption easier for UK legal teams is helping firms make that exploration time count
The clearest guidance for practitioners in the interim remains straightforward: know where you’ve used AI, use authoritative sources rather than open web tools, verify what it produces, and do not cite cases that don’t exist. The SRA will not ignore this indefinitely.
The strategic window is open — for now
The Thomson Reuters Institute has outlined four plausible futures for law firms:
- The tech-led disruptor — agentic AI does the heavy lifting; subscription and self-serve models replace billable hours
- The elite boutique — exceptional human expertise amplified by AI; the model lives or dies on protecting the firm’s IP
- The integrated powerhouse — human methodology systematised and delivered at scale via AI
- The traditional preserver — described candidly as a delay, not a strategy
Most firms will find themselves gravitating towards one of the first three. The fourth offers the comfort of familiarity but demands an honest question: how long does that runway actually last?
What was striking about British Legal Tech Forum 2026 wasn’t panic or hype — it was the growing clarity that the choices firms make in the next 12 to 18 months will define their competitive position for a decade. That means picking a strategic identity and committing to it. It means treating culture change with the same budget and seriousness as technology licensing. It means moving beyond operational tick-boxes to measure what genuinely matters — the quality of legal outcomes, the strength of client relationships, and the ability to attract and retain the next generation of talent.
What the next 18 months mean
The legal profession has navigated every previous wave of change — from digitisation to globalisation — not by waiting for certainty, but by acting with intent before the path was fully clear. The firms that thrive in the AI era will be those that do the same: not the loudest adopters, not the most cautious sceptics, but the ones who asked the harder questions early — and started building the answers.
Find out how CoCounsel Legal UK can help your firm take that next step.
CoCounsel Legal UK
Setting a new standard for AI-driven legal work, offering advanced AI, trusted content, and seamless integrations, all in one place.
Learn more ↗