How AI, data and a £55.5bn UK legal market could reshape origination, partnership economics and client trust

For generations, the economics of the law-firm partnership rested on a recognisable chain. A partner built a relationship with a general counsel, board member or entrepreneur; trust transferred through that relationship, instructions followed and origination credit helped determine status and compensation. That model remains powerful. But the buyer of legal services is changing faster than the partnership structure that serves it.

According to the Law Gazette, citing IRN Legal Reports, turnover across UK legal activities reached £55.5 billion in 2025, up 6.1% at current prices and around 50% since 2020. Business and commercial legal advice, including commercial property, accounted for 51% of the market. Even a modest change in how corporate clients allocate legal work could therefore redirect billions of pounds of annual spend. The temptation is to leap from that fact to a dramatic conclusion: AI agents will choose law firms, personal relationships will become irrelevant and the traditional rainmaker could become the first casualty. That is possible in parts of the market, but it is not yet established. There is limited public evidence that autonomous agents are routinely selecting and instructing outside counsel without human oversight. What is much better evidenced is the infrastructure that could make such a shift plausible.

Corporate legal departments are adopting AI rapidly and are increasingly asking their advisers to do the same. The 2026 State of the UK Legal Market from Thomson Reuters found that 78% of corporate legal respondents believed AI should be applied to legal work within law firms, whilst 69% expected generative AI to become central to their organisation’s workflows within two years. That does not mean an algorithm has replaced the general counsel. What is does mean is that the client is becoming more technologically capable and may expect outside counsel to provide measurable evidence of efficiency as well as legal expertise. The same direction is visible in legal operations. Thomson Reuters’ 2025 Legal Department Operations Index encourages departments to collect and use data when selecting and evaluating outside counsel, including information on technology use and performance. This matters because AI is particularly good at comparing structured information. A future procurement tool does not need to “dislike” relationships; it merely needs enough reliable data to ask whether a different firm can deliver comparable expertise faster, more predictably or at lower cost. That is a subtler disruption than the disappearance of reputation. Reputation remains valuable because legal work is not a commodity. In major litigation, investigations, restructuring, M&A and board-level crises, judgement, confidentiality, conflicts, strategic instinct and the credibility of individual lawyers matter enormously. The change is that reputation may increasingly need to co-exist with harder evidence rather than substitute for it.

From “who do we know?” to “what can we verify?”

An AI-assisted buyer could theoretically compare experience by sector, matter type, jurisdiction, staffing model, cycle time, budget variance, complaints, conflicts, technology capability and client-defined outcomes. Some of those measures are already available internally to sophisticated legal departments; others would be difficult to compare fairly. A complex cross-border dispute cannot be reduced to a league table in the same way as a routine contract review. This is where the original “machine-readable law firm” argument becomes useful. Firms may increasingly need to structure information about what they do, how they price it and how work is progressing so that a client’s systems can interrogate it. That could mean richer client portals, standardised matter taxonomies and controlled data feeds rather than a public API exposing confidential performance data. The objective is not to publish every matter - it is to make relevant evidence available to authorised clients in a form their systems can understand.

If more instructions are influenced by data-driven procurement, the economics of origination could change. Today, many partnerships reward the partner who “owns” the relationship or brings the client through the door. In a more digital buying process, that door may be opened by a combination of brand, client data, pricing, sector credentials, technology and measurable delivery. It becomes harder to attribute the mandate to one individual. That does not mean origination credit disappears. A senior partner may still create the trust that persuades a client to include the firm in the first place, interpret the client’s unspoken concerns and take responsibility when the matter becomes difficult. But partnership economics may gradually place more value on people who build scalable capabilities: legal products, data taxonomies, knowledge systems, AI governance, client integration and teams that deliver consistently across matters. The question for partnerships is therefore not “relationships or technology?” It is whether their reward systems recognise both. A firm that compensates only personal origination may under-invest in the shared infrastructure that increasingly supports client acquisition and delivery. A firm that treats technology as a substitute for judgement risks the opposite mistake.

AI is entering regulated legal delivery - but humans remain accountable

The UK regulatory direction is equally nuanced. In May 2025, the Solicitors Regulation Authority authorised Garfield.Law, (described by the regulator as the first purely AI-based law firm it had authorised to provide regulated legal services), focused on debt recovery. The example demonstrates that AI can move beyond internal productivity tools into regulated service delivery. But it does not remove professional responsibility. The SRA’s 2026 warning notice on misuse of AI states that the use of AI does not diminish or transfer a solicitor’s professional responsibilities. The Law Society’s October 2026 practice note similarly stresses human verification, professional judgement, confidentiality, privilege, data protection and governance. If clients increasingly use AI to buy legal services, those controls may themselves become part of what the client measures. There are three changes worth considering:

· First, firms may need better internal data before they need external AI. Matter descriptions, outcomes, budgets, staffing, cycle times and client feedback are often stored inconsistently. If a firm cannot interrogate its own experience, it will struggle to provide reliable evidence to a client or procurement system. The objective should be consistency, not indiscriminate transparency.

· Second, reporting is likely to become less static. Quarterly PDFs will not disappear, but sophisticated clients may increasingly expect dashboards, live spend information, matter status and agreed KPIs. That creates a commercial opportunity for firms that can make their service easier to monitor without compromising confidentiality or professional independence.

· Third, the public role of senior lawyers may evolve. Thought leadership, conferences and personal networks still matter because humans remain involved in high-value legal buying. But partners may increasingly need to demonstrate a second form of credibility: that the firm can use AI safely, explain how decisions are supervised, disclose where automation is used and show that efficiency gains are being shared with the client.

The UK legal market is not on the verge of handing £55.5 billion of annual revenue to autonomous procurement agents. Relationships, reputation and professional judgement remain embedded in how legal risk is bought and managed. Nor is every legal outcome capable of being scored objectively. The danger of pretending otherwise is that difficult legal work becomes reduced to metrics that reward speed over quality or apparent success over prudent advice. Yet the opposite assumption may be equally risky: that because law is a relationship business, the purchasing process will remain fundamentally human. Corporate legal teams are already using more AI, bringing more work in-house and asking external firms to demonstrate value, efficiency and technological capability. The next step need not be an autonomous agent with power to instruct counsel; it may simply be an AI-enabled general counsel with far better information before deciding whom to call. If that happens, the rainmaker does not vanish - the role changes. Personal trust may still win the most difficult mandate but data may determine who reaches the shortlist, what price the client regards as credible and whether the relationship survives the next review. In that world, the most valuable partner may not be the person with the largest contacts book or the best dashboard. It may be the lawyer who can combine reputation, judgement and accountability with a service whose value can increasingly be verified. It may also mean rainmakers are more mobile and more flexible in the way they practice; with AI providing the back-office infrastructure and support of a large firm at a fraction of the cost, why would a rainmaker want to profit-share anymore? Why not practice alone?

For law-firm leaders, the more useful question is not therefore whether AI will kill the partnership. It is whether a partnership designed around human origination can evolve quickly enough for a market in which trust is still human, but more of the evidence behind that trust is becoming machine-readable. However, if a law firm cannot agentically communicate with AI-powered agents, how much business are they missing out on? Given Accenture estimates that, by 2030, AI agents could handle more than 30% of online commerce, amounting to nearly US$3.1 trillion in transactions, that could well equate to a great deal of potential legal work mandates including disputes and litigation.

So, is your lawyer rainmaker ready?

This article first appeared in Digital Bytes (10th of November, 2026), a weekly newsletter by Jonny Fry of Team Blockchain.

How could AI change the way clients choose law firms and how legal partnerships create value?