How do law firms actually use AI in 2026?

Law firms in 2026 use AI primarily for client intake automation, document drafting, compliance checking, and client communications. 62% of UK law firms now use at least one AI application according to the Law Society’s Digital Benchmark, but only 29% have implemented AI as part of a coherent firm-wide strategy. The gap between experimentation and strategic deployment defines the current landscape.

Short answer: Intake, documents, compliance, and client comms are the four core use cases. 62% of UK firms use AI somewhere, but only 29% have a real strategy behind it.

Why this question matters now

The gap between “using AI” and “using AI well” is widening. Most firms have someone using ChatGPT for research or drafting. Fewer have thought about how AI integrates with their practice management system, their compliance obligations, and their business model.

This matters because AI adoption is no longer optional for competitive firms. Corporate clients on panel arrangements are asking about AI capabilities. Legal aid providers are under pressure to do more with less. Mid-market firms are watching competitors announce AI initiatives and wondering whether they are falling behind.

The reality is more nuanced than either the hype or the scepticism suggests. AI is not replacing lawyers. It is not a silver bullet. But firms that deploy it thoughtfully against the right use cases are seeing measurable improvements in efficiency, client satisfaction, and profitability. Understanding what “actually using AI” looks like in practice, rather than in vendor presentations, is the first step.

Client intake and triage

This is the highest-impact, lowest-risk AI use case for most firms. A well-built intake system qualifies enquiries before a solicitor gets involved, capturing relevant details through intelligent forms or conversational interfaces, routing matters to the right team, and providing immediate acknowledgement to potential clients.

The ROI is compelling because intake is high-volume, largely repetitive, and currently handled by expensive solicitor time. A family law firm receiving 200 enquiries per month might have a solicitor spending 15 minutes on each initial call, many of which are outside the firm’s practice area or not viable matters. That is 50 hours of solicitor time per month on triage.

An AI intake system handles the initial qualification, collects structured information, and presents solicitors with pre-qualified, well-documented leads. The solicitor’s first interaction with a potential client is informed and efficient rather than exploratory.

In the US, firms are seeing similar gains. Immigration, personal injury, and employment practices with high enquiry volumes benefit most. The key is ensuring the intake system is properly configured for the jurisdiction, as qualification criteria and regulatory requirements differ by state.

Document drafting and assembly

AI-assisted document drafting has moved beyond novelty into daily use at many firms. The practical applications include first drafts of engagement letters and client care correspondence, standard contract clauses with firm-specific variations, witness statements from interview notes, court form population from case data, and routine correspondence such as completion letters in conveyancing.

The important distinction is between generation and automation. AI generates new text based on context. Document automation populates templates with data. Both are useful, and the best implementations combine them. A conveyancing completion letter might use automation for the structured elements (names, dates, property details) and generation for the contextual narrative.

Human review remains essential. No responsible firm sends AI-generated documents to clients or courts without solicitor review. The efficiency gain comes from reviewing and refining a competent first draft rather than starting from a blank page. Depending on document complexity, this saves 40 to 70% of drafting time.

Tools in this space include Harvey for general legal drafting, Luminance for contract-specific work, and custom systems built on large language model APIs for firm-specific document types.

Compliance checking and risk management

AI compliance tools monitor regulatory changes, flag potential conflicts, and check documents against regulatory requirements. In conveyancing, AI systems verify that searches are complete and consistent. In employment law, they check that settlement agreements comply with current statutory requirements.

The SRA’s 2025 Technology Guidance makes clear that firms must understand and take responsibility for AI-assisted compliance decisions. The AI does not replace the compliance officer; it makes the compliance officer more effective by catching issues that manual review might miss.

In the US, compliance applications vary by practice area. Immigration firms use AI to track visa bulletin changes and filing deadlines. Securities practices use it to monitor regulatory updates from the SEC and FINRA. The common thread is that AI handles the monitoring and flagging, while humans make the judgment calls.

Anti-money laundering (AML) checks are another growing use case. AI systems can screen clients against sanctions lists, PEPs databases, and adverse media more quickly and consistently than manual processes, though human review of flagged cases remains mandatory.

Client communications and updates

This is the use case firms are most cautious about, and rightly so. AI-generated client communications must be accurate, appropriately toned, and compliant with professional obligations. But when done well, automated client updates dramatically improve the client experience.

Practical applications include automated case progress updates (e.g., “Your search results have been received and are being reviewed”), AI-drafted response templates for common client queries, meeting summary emails generated from notes, and deadline reminders with context.

The key is keeping a human in the loop for substantive communications while letting AI handle informational updates. A client who receives a weekly progress update, even an automated one, is significantly less likely to call the firm asking what is happening. This reduces inbound call volume and improves client satisfaction simultaneously.

What we have seen at Formulaic

The pattern across our clients is consistent: intake is the gateway use case. When we built the employment law intake system for Calder & Reid, the immediate impact was a 70% reduction in unqualified calls and £78,000 per year in saved solicitor time. But the secondary effects were equally important. Solicitors reported higher job satisfaction because they were spending more time on substantive legal work and less on telephone triage. Client conversion rates improved because qualified leads were handled faster and with better information.

From 30 production AI systems shipped across 6 clients, our data shows that firms get the fastest ROI from intake automation (weeks to payback), the highest total savings from document drafting at scale, and the best client satisfaction improvements from automated communications. The firms that struggle are the ones trying to do all four simultaneously. Start with intake, prove the value, then expand.