Artificial intelligence and the business of conveyancing
Labour is the largest cost in almost every conveyancing practice. Artificial intelligence is now automating the administrative layer where most of that labour is spent. A reduction of around thirty per cent in labour cost per transaction is no longer a futurist's slide. It is a planning assumption.
The firms that treat it as one will set the price. The rest will spend the next five years explaining why they cannot match it.
The conveyancing profession has spent the past two years discussing artificial intelligence in the abstract. The conversation has been about pilots, about caution, and about whether the technology is genuinely ready. That conversation is now behind the curve. The serious question for any firm is no longer whether artificial intelligence will change the economics of conveyancing. It is how far, how fast, and whether your firm will be the one setting the new price or the one conceding it.
This article makes a direct claim. Within five years, a well run conveyancing practice should be able to reduce its labour cost per transaction by around thirty per cent through the disciplined adoption of artificial intelligence. That is not a reduction achieved by cutting corners on legal work. It is a reduction achieved by removing the manual administrative burden that currently consumes the majority of fee earner and support staff time, and by reserving qualified human judgment for the work that genuinely requires it.
The claim deserves scrutiny, and the sections that follow provide it, together with a five year business plan for realising the saving without compromising the firm's regulatory standing. That last point is not a footnote. It is the difference between a saving and a deferred liability.
Where the Money Actually Goes
Begin with the cost structure, because the rest of the argument follows from it. In most conveyancing practices, staff costs are the dominant line in the profit and loss account, frequently somewhere between half and two thirds of total expenditure. If you want to reduce the cost of producing a conveyancing transaction, you have to address labour. There is nowhere else of comparable size to go.
The instinctive professional response is defensive. Conveyancing is skilled legal work, the argument runs, and you cannot simply automate skilled legal work without harming clients. That argument is partly true and largely beside the point.
The uncomfortable reality, which the profession has been reluctant to state plainly, is that a large proportion of what happens inside a conveyancing file is not legal work at all. It is administration. Opening the file. Sending the client care letter. Requesting identification. Chasing the other side. Acknowledging documents. Logging searches when they return. Producing status updates for anxious buyers and sellers. Populating standard forms. Producing the first draft of standard documents. Much of this does not require qualified legal judgment where the matter is routine, provided that exceptions are identified and escalated. All of it currently consumes the time of people the firm is paying.
This is not a controversial observation among those who design conveyancing systems. Commentary across the sector now describes the same pattern: the legal expertise inside firms is strong, but the surrounding administration remains heavily manual, and experienced staff lose a significant part of every day to repetitive status requests and basic file administration.
A large proportion of what happens inside a conveyancing file is not legal work at all. It is administration. And administration is precisely what artificial intelligence now does well.
The Evidence Is No Longer Speculative
This used to be a theoretical argument. It is not any longer. The data from the past eighteen months has moved the question from possibility to measurement, although the figures should be read as reported gains that depend on disciplined implementation rather than guaranteed outcomes that arrive simply because a tool has been purchased.
In the United Kingdom, Clio's research found that forty three per cent of solicitors report that artificial intelligence has already improved their productivity, with a further benefit reported in wellbeing and work life balance, and that a third of firms intend to invest more than one hundred thousand pounds in technology over the coming year. The direction of travel is not in doubt.
The Thomson Reuters and Georgetown report on the state of the legal market for 2026, which reflects the wider United States and global legal market rather than the United Kingdom conveyancing sector specifically, recorded that firm technology budgets rose sharply against pre generative levels, and that firms with a formal artificial intelligence strategy were close to four times more likely to realise critical benefits than firms without one. That evidence is contextual rather than direct proof for conveyancing, but the lesson it carries is not sector specific. Strategy is the variable that separates the firms that capture the saving from the firms that merely spend the money.
At the level of the individual professional, vendor commissioned research points the same way, although it should be read as exactly that. An independent adoption study of one leading legal platform, surveying around forty firms and in house teams, reported monthly time savings of roughly sixteen hours for regular users, rising to around thirty seven hours for the heaviest users, with the large majority of participants agreeing that legal work was being executed faster. Those figures are self reported and were commissioned by the vendor, so they are best treated as indicative rather than definitive. Separate research into the future of the professions has estimated that legal professionals expect to free up close to two hundred and forty hours each year through artificial intelligence adoption.
Apply those figures to conveyancing specifically. The Law Society has confirmed that artificial intelligence is already being used across the conveyancing lifecycle for title checks, search report review, examination of mortgage offers, document review, drafting of standard contracts, and workflow automation. These are not exotic future applications. They are the administrative core of the transaction, and they are being automated now.
The Competitive Threat Is Already in the Market
The most important development is not a statistic. It is a competitor. An artificial intelligence first conveyancing firm, operating under the working name Keith and seeking authorisation from the Council for Licensed Conveyancers, has stated publicly that it intends to be the most automated firm in conveyancing and expects to reduce transaction times by around seventy per cent. Its stated pricing intention sits at the lower end of the market.
Whether that particular venture succeeds is not the point. The point is that the model is funded, reportedly to the order of two million pounds, and is being developed with Council for Licensed Conveyancers authorisation being sought, rather than already proven in operation. When a firm constructed from the ground up around automation sets out to complete transactions faster and price at the bottom of the market, the pressure on traditional firms is not gradual. It is structural. Either you reduce your own cost of production to compete, or you concede the volume end of the market and retreat to work where price is less decisive.
A reduction of around thirty per cent in labour cost is, in that light, not an ambition. It is closer to the entry fee for remaining competitive in volume conveyancing over the next five years.
What a Thirty Per Cent Reduction Actually Means
Be precise about the claim, because imprecision here invites both false comfort and false alarm.
The reduction is a reduction in labour cost per transaction. It is not necessarily a thirty per cent cut to the payroll. Those are different things, and the difference matters enormously for how the firm plans and for how it treats its people.
A firm can realise the saving in two broad ways. The first is capacity. The same number of staff handle substantially more matters, so the labour cost attributable to each transaction falls even though the wage bill does not. The second is headcount. As routine administrative roles are automated, the firm employs fewer people to handle the same volume. Most firms will use a combination, and the balance they choose is a matter of strategy and circumstance.
AN ILLUSTRATIVE MODEL The figure is not plucked from the air. It follows from the cost structure. Assume that administrative and process work accounts for between fifty five and sixty five per cent of the labour on a routine file. Assume further that artificial intelligence and workflow redesign remove between forty and fifty per cent of that administrative burden. The resulting reduction in file labour lands in the region of twenty two to thirty two per cent, before any further effect from redeployment or the additional supervision the redesigned process requires. The thirty per cent sits within that range for a firm that executes well on routine, process led work. It is a planning assumption for such a firm. It is not a guaranteed market wide outcome, and a practice weighted towards complex or non routine matters should expect a lower figure. |
The honest position, which the profession owes its own people, is this. Some firms will realise the saving primarily through increased capacity and redeployment. Others will make structural headcount changes over time. The responsible distinction is not which route a firm takes, but whether that transition is planned, transparent, and governed. The firm that plans deliberately, retrains where it can, redeploys people towards work that genuinely needs human judgment, and is candid throughout protects both its people and its reputation. The firm that denies the change until it arrives makes abrupt decisions under pressure and damages both.
The Five Year Business Plan
A saving of this scale is not captured by enthusiasm. It is captured by sequence. The following plan moves deliberately from foundations to advantage, and at every stage it treats governance as a condition of progress rather than an obstacle to it.
Year One: Foundations, Governance, and an Honest Audit
Before a single efficiency is captured, the firm needs to know two things: what its people are actually doing with their time, and what its rules permit. Skip this stage and every later number is a guess.
Begin with a candid audit of how fee earner and support staff time is spent across a representative sample of matters. Distinguish, file by file, between time spent on administration and time spent on genuine legal judgment. This is the baseline against which every later saving is measured, and without it the firm will never be able to demonstrate that the thirty per cent has actually been achieved.
In parallel, establish governance. The Upper Tribunal has already made clear that placing client information into public or consumer artificial intelligence tools is a data breach, a breach of confidentiality, and a waiver of privilege. The Tribunal used the phrase open source colloquially, but the distinction that matters is not one of model licensing. It is the difference between uncontrolled public tools and secure, contracted, closed environments. A firm that pursues efficiency without a governance framework is not saving money. It is accumulating regulatory risk that will eventually cost far more than the saving it chased.
REQUIRED ACTIONS: YEAR ONE • Conduct a documented time audit across a representative sample of matters, separating administrative time from time spent on legal judgment. • Adopt a written artificial intelligence acceptable use policy, signed off at partner or board level, distinguishing secure, contracted, closed environments that may be approved from uncontrolled public or consumer tools that must not be used with client data. • Confirm that any tool under consideration keeps client data within a controlled environment, supported by an appropriate data processing agreement. • Define the supervision and verification framework before deployment, not after. |
Year Two: Targeted Deployment in the Administrative Layer
Deploy where the return is clearest and the risk is lowest. The administrative layer identified in the year one audit is the obvious target: file opening, identity verification, client status communication, acknowledgement and logging of documents, summarising of search results and title information, and drafting of standard contractual documents.
Measure relentlessly against the year one baseline. The objective at this stage is not transformation. It is proof, in numbers the firm can trust, that time is being released and cost per transaction is beginning to fall.
CAUTION: EFFICIENCY MUST NOT OUTRUN SUPERVISION A tool that summarises a title or a search result produces a draft, not a conclusion. Every output that informs advice to a client or a filing must be checked by a qualified person. Speed gained at the cost of an unchecked output is not a saving. It is the precise failure that has already drawn regulatory attention across the profession. |
Year Three: Redesigning Roles, Not Just Adding Tools
The error most firms make is bolting artificial intelligence onto an unchanged process and then wondering why the promised saving never fully arrives. The larger gains come from redesigning the workflow around the tools, and from redesigning the roles of the people who operate it.
Commentary within the wider professional research, including the Wolters Kluwer Future Ready Lawyer material, has described an eighty twenty reversal, in which fee earners come to spend the majority of their time on advice and judgment rather than on administration. That is offered as a direction of travel rather than a settled empirical finding, but it is the shape the year three firm should be aiming for. The administrative paralegal role changes character, becoming a role focused on overseeing automated processes and handling the exceptions that the system flags. Supervision does not diminish in this model. It intensifies, because the qualified professional is now responsible for the output of a process rather than for keying every step of it by hand.
Year Four: Repricing and Restructuring
By year four the saving should be real and measured. The strategic decision is what to do with it. There are three broad options, and the firm should choose deliberately rather than by default.
• Pass the saving to clients as lower fixed fees in order to win volume and defend market share against automated entrants.
• Retain the saving as margin to strengthen the firm's financial resilience, accepting that this position is harder to hold as competitors reprice.
• Reinvest the saving in capability, capacity, or the higher value advisory work that automation does not threaten.
Whichever path is chosen, the billing model must be confronted honestly. Conveyancing already largely operates on fixed fees, which makes it more exposed to price competition than practice areas that still bill by the hour. When a competitor can produce the same transaction at a structurally lower cost, a firm that has not reduced its own cost base has no answer to give the client who asks why it charges more.
Year Five: Consolidation and a Defensible Advantage
By year five the firm should be operating at a labour cost per transaction materially below where it started, with the reduction approaching thirty per cent in a practice that has executed the plan with discipline. The advantage is defensible precisely because it does not rest on a single tool that a competitor can simply buy. It rests on a redesigned process, retrained people, and embedded governance, which together are far harder to replicate.
The work does not stop. The tools, their accuracy, and the regulatory framework around them all continue to move. Continuous monitoring becomes a permanent function rather than a project, and the firm that built the habit early is the firm best placed to absorb the next change without disruption.
The plan in summary:
Horizon | Primary focus | Target outcome |
Year one | Honest time audit and governance foundations. Establish the artificial intelligence acceptable use policy and supervision framework before any tool is deployed. | A measured baseline of where staff time is spent, and a compliant framework within which efficiency can be pursued. |
Year two | Targeted deployment in the administrative layer: file opening, identity verification, status communication, search and title summarising, standard drafting. | Early, measured efficiency gains against the year one baseline, captured without regulatory exposure. |
Year three | Redesign roles and workflow around the tools rather than bolting tools onto an unchanged process. Reposition fee earners towards judgment and advice. | A process built for automation, with supervision intensified rather than diluted. |
Year four | Reprice and restructure. Decide deliberately whether the saving is passed to clients to win volume, retained as margin, or reinvested. | A labour cost per transaction materially below the year one baseline, with pricing aligned to the new cost base. |
Year five | Consolidate the advantage. Embed continuous monitoring of tools, accuracy, and regulatory developments. | A reduction in labour cost per transaction of around thirty per cent, defensible because it rests on redesigned process and embedded governance. |
The Risk and Compliance Dimension Is Not an Afterthought
Cost reduction that creates regulatory exposure is not a saving. It is a deferred liability, and it tends to surface at the worst possible moment. The firms that capture the saving safely are the firms that treat the following obligations as part of the design, not as a compliance review bolted on at the end. The SRA's own guidance on the use of artificial intelligence and technology points the same way, identifying leadership oversight, risk and impact assessments, written policies, training, and ongoing monitoring as the expected components of responsible adoption.
Supervision and accountability. Paragraph 3.5 of the SRA Code makes those who supervise or manage others accountable for the work carried out through them. The Upper Tribunal has put the point beyond argument: it does not matter how an error comes about, because the qualified professional with conduct of the matter is responsible for the accuracy of what is filed. Automating the production of a document does not transfer that responsibility to the software.
Competence. Paragraphs 3.3 and 7.1 require solicitors to maintain their competence and to keep up to date with the law and regulation governing their work. In a firm deploying artificial intelligence, competence now includes understanding what the tools can and cannot do, and where their outputs must be checked.
Confidentiality, privilege, and data. Client information placed into a public or consumer tool is placed into the public domain. That is a data breach, a waiver of privilege, and a breach of the duty of confidentiality, and it may require assessment for notification to the Information Commissioner's Office and for regulatory reporting, the obligation in each case depending on the threshold, the seriousness, and the risk involved. The efficiency case for artificial intelligence collapses entirely if it is pursued outside a secure, contracted environment.
Verification of output. Artificial intelligence generates plausible text, and plausible text is not the same as accurate text. Fictitious case citations are the notorious example, but in conveyancing the risk is quieter and just as serious: a summarised title, an interpreted search result, or a drafted clause that is subtly wrong. Verification by a qualified person must be a mandatory checkpoint in the workflow, not an optional courtesy.
Anti money laundering. Automation can streamline the gathering of identity and source of funds material, but it does not discharge the firm's obligations under the regulations. A bank statement is a starting point, not a conclusion. Proof of balance is not proof of provenance, and the assessment of where funds have genuinely come from remains a matter of human judgment that no current tool can be permitted to make on the firm's behalf.
There is a constructive point to set against these warnings, and it deserves equal weight. A well designed automated process can strengthen compliance rather than weaken it. Consistency improves when every file is opened the same way and every required step is prompted rather than remembered. The audit trail improves when actions are logged automatically. Used deliberately, the same systems that reduce labour cost can also reduce the risk of the human error and inconsistency that compliance functions spend their time chasing.
HIGH RISK INDICATORS: STOP AND ESCALATE • Any use of a public or consumer artificial intelligence tool with client information, privileged material, or confidential data. • Artificial intelligence output reaching a client or a third party without verification by a qualified person. • Source of funds or wider anti money laundering conclusions being treated as settled on the strength of an automated check alone. • Headcount reductions implemented before the supervision framework has been redesigned to cover the automated process. • Deployment proceeding without a written, partner level acceptable use policy in place. |
The Honest Conclusion
The thirty per cent is achievable. The evidence from the wider legal market, the specific applications already live across the conveyancing lifecycle, and the arrival of automated competitors all point the same way. A firm that audits its time honestly, governs its tools properly, redesigns its process rather than decorating it, and reprices deliberately can reach that figure within five years.
Whether it is achieved safely is a separate question, and that is where the risk and compliance function earns its place at the planning table rather than at the end of the corridor. The temptation, under competitive pressure, will be to chase the saving and treat the regulatory framework as a brake. That instinct is precisely backwards. The framework is what makes the saving durable, because a saving that is later unwound by a data breach, a misfiled document, or an anti money laundering failure was never a saving at all.
The profession has a choice that mirrors the one it faces on automation more broadly. It can insist that conveyancing is too complex to change and wait for the change to be imposed by competitors who have already decided otherwise. Or it can plan the transition on its own terms, capture the efficiency deliberately, protect its clients and its people through the change, and emerge as the kind of firm that sets the price rather than apologises for it.
The technology will not wait for the decision. The firms that begin the five year plan this year will be the ones still setting terms when it is finished.
References
Sources are listed in the order in which the underlying material appears in the article. Figures drawn from market research and vendor case studies represent reported outcomes and should be validated against the firm's own environment.
1. Clio, How UK Law Firms Are Embracing AI and Technology in 2026. Source for the finding that forty three per cent of UK solicitors report improved productivity from artificial intelligence, the reported wellbeing benefit, and the finding that a third of firms intend to invest more than one hundred thousand pounds in technology over the coming year. Directly relevant United Kingdom evidence. Accessed June 2026. https://www.clio.com/uk/blog/ai-technology-trends/
2. Thomson Reuters Institute and Georgetown Law, 2026 Report on the State of the Legal Market (analysis). Contextual evidence drawn from the wider United States and global legal market rather than United Kingdom conveyancing specifically. Source for the sharp rise in firm technology budgets against pre generative levels, and for the finding that firms with a formal artificial intelligence strategy are close to four times more likely to realise critical benefits. Accessed June 2026. https://www.thomsonreuters.com/en-us/posts/legal/legal-market-report-2026-analysis-ai-bubble/
3. Thomson Reuters, The New Economics of AI Powered Legal Services. Source for the Future of Professionals estimate that legal professionals expect to free up close to two hundred and forty hours each year through artificial intelligence adoption, and for the wider analysis of value and billing. Accessed June 2026. https://legal.thomsonreuters.com/blog/the-new-economics-of-ai-powered-legal-services-how-smart-law-firms-are-redefining-profit/
4. Legal IT Insider, A deeper dive into the RSGI and Harvey adoption report. Source for the per professional time savings cited, drawn from an adoption study of around forty firms and in house teams reporting monthly savings of roughly sixteen hours for standard users and around thirty seven hours for power users, with the large majority agreeing legal work was being executed faster. The research was commissioned by the vendor and the figures are self reported, so they are treated as indicative rather than definitive. Accessed June 2026. https://legaltechnology.com/2025/12/02/the-impact-of-legal-ai-a-deeper-dive-into-the-rsgi-harvey-adoption-report/
5. The Law Society, Conveyancing and AI: finding the right approach for your firm. Directly relevant source confirming that artificial intelligence is already used across the conveyancing lifecycle, including title checks, search report review, examination of mortgage offers, document review, drafting of standard contracts, correspondence, and client update automation. Accessed June 2026. https://www.lawsociety.org.uk/topics/property/conveyancing-and-ai-finding-the-right-approach-for-your-firm
6. Legal Futures, Meet Keith: the AI first law firm looking to transform conveyancing. Source for the automation first conveyancing venture, reportedly funded to the order of two million pounds, with Council for Licensed Conveyancers authorisation being sought, an ambition of around eighty per cent automation, a stated expectation of reducing transaction times by around seventy per cent, and a lower end pricing intention. Accessed June 2026. https://www.legalfutures.co.uk/latest-news/meet-keith-the-ai-first-law-firm-looking-to-transform-conveyancing
7. Solicitors Regulation Authority, SRA Code of Conduct for Solicitors, RELs, RFLs and RSLs. Source for the supervision and competence obligations relied upon, including paragraph 3.3 on maintaining competence, paragraph 3.5 making those who manage or supervise accountable for work carried out through them, and paragraph 7.1 on keeping up to date with the law and regulation governing the work. Accessed June 2026. https://www.sra.org.uk/solicitors/standards-regulations/code-conduct-solicitors/
8. Solicitors Regulation Authority, Compliance tips for solicitors regarding the use of AI and technology. Source for the governance components identified as expected of responsible adoption, including leadership oversight, risk and impact assessments, written policies, training, and ongoing monitoring. Accessed June 2026. https://rules.sra.org.uk/solicitors/resources/innovate/compliance-tips-for-solicitors/
9. Birketts, Using AI and the risk of waiving legal privilege. Source clarifying that the Upper Tribunal used the phrase open source colloquially, and that the operative distinction is between uncontrolled public tools and secure, contracted, closed environments rather than a question of model licensing. Accessed June 2026. https://www.birketts.co.uk/legal-update/using-ai-and-the-risk-of-waiving-legal-privilege/
10. Wolters Kluwer, Future Ready Lawyer 2026. Commentary within a broader global report. Source for the eighty twenty reversal, in which lawyers move from routine work towards higher value strategic advice. Cited as a direction of travel rather than a settled empirical finding. Accessed June 2026. https://www.wolterskluwer.com/en/know/future-ready-lawyer-2026
11. Nexform AI, Conveyancing Automation for UK Law Firms. Source for the description of the heavily manual administrative layer in conveyancing and the loss of fee earner and support staff capacity to repetitive status requests and basic file administration. Accessed June 2026. https://nexformai.co.uk/conveyancing-automation-for-uk-law-firms-a-smarter-way-to-keep-buyers-and-sellers-updated/
12. Legal Futures, The 2026 conveyancing reset. Context for the shift towards front loaded information, digital identity, and workflow ready artificial intelligence in conveyancing, including the introduction of the updated Law Society transaction forms for accredited firms. Accessed June 2026. https://www.legalfutures.co.uk/associate-news/the-2026-conveyancing-reset-why-upfront-data-digital-id-and-ai-optimised-workflows-will-define-the-next-era-of-property-transactions
This article is intended as general guidance and does not constitute legal advice. The figures cited reflect reported research and vendor case studies and should be independently validated before being relied upon. If you would like to discuss artificial intelligence governance, adoption strategy, or the compliance framework around it for your practice, please get in touch.
Comments
Post a Comment