A solicitor running a three-partner firm recently described her situation. She had renewed her AI drafting tool subscription. The tool now generates a first-draft employment contract in twelve minutes. Before it, the same task took two to three hours. Her fee to the client had not changed.
That is where the economics break quietly. She is billing fewer hours for the same outcome. She has not yet worked out how to price for the value she delivers, the contract itself, or the professional liability she carries if something goes wrong. Her pricing model is still built for a world where time and output moved together.
What is value-and-risk pricing for AI-enabled legal services?
Value-and-risk pricing means setting your fee based on what the outcome is worth to the client and the level of professional responsibility you are taking on, rather than the hours you expect to spend. For AI-enabled legal work, this matters because AI reduces the time required for many tasks without reducing the legal judgement, accountability, or professional liability involved in delivering them.
The fee for drafting a shareholder agreement is not primarily compensation for four hours of typing. It reflects your expertise, your indemnity exposure if the agreement is later contested, and the commercial value the document protects for the client. AI can accelerate the drafting without changing any of those three things. A pricing model anchored purely to time will compress your margins as AI usage grows, because the hours shrink while the accountability does not.
Value-based and risk-tiered fees are not new in legal practice. Linklaters has publicly reported that around 60% of its mandates now use non-hourly pricing, partly supported by AI-driven process improvement and matter analytics. For small and mid-sized firms, the shift has been slower, but AI is making it more viable by creating the cost predictability you need to quote a fixed fee with confidence.
Why does the AI productivity squeeze make this urgent now?
If you have adopted AI tools and kept hourly billing, you face a two-sided squeeze. AI reduces the time your people spend on routine work, cutting billable hours. Meanwhile clients are seeing faster turnaround and starting to ask why the fee is unchanged. Research suggests 74% of hourly-billed tasks at small and mid-sized UK firms are automatable with AI and automation tools.
UK lawyers are expected to save around 140 hours per lawyer per year by 2026, equivalent to roughly £12,000 in time savings per lawyer annually at current employment costs. Across the sector, the projected productivity gain is £2.4 billion. If your pricing model treats those saved hours as write-offs rather than as recoverable margin, you pass the entire benefit to clients without a strategy for capturing any of it yourself.
Client expectations are also shifting. Around 70% of clients surveyed say they prefer or are indifferent to AI-powered legal services where they gain transparency, speed, and fixed-fee pricing. Small firms using AI-backed chatbots and document generators alongside fixed-fee offerings report around 30% higher client retention rates. The business case for changing how you price is clearer now than it was eighteen months ago.
Where does this pricing approach actually apply?
Value-and-risk pricing works best when work can be categorised by its risk profile and what the outcome is worth to the client. In practice, this creates three broad categories for small UK firms: standard work handled largely by AI with light review, bespoke work combining AI efficiency with experienced solicitors, and high-stakes matters where hybrid fee arrangements still make more sense.
The lowest-risk category covers work where scope is tight and AI-accelerated output is reliable with supervision. Standard contracts, NDAs, routine compliance packs, and employment handbook reviews sit here. Fixed fees work for this category because AI brings predictability to your cost side, and clients value the certainty on their side.
A middle tier covers work with more complexity or bespoke elements, where an experienced solicitor is directing the AI rather than simply reviewing its output. Fixed fees with defined assumptions and explicit change-control clauses work here. Any work outside the stated scope should be priced separately, and the engagement letter needs to reflect that clearly.
High-stakes matters, contested disputes, complex M&A, high-value employment cases, remain better suited to hourly, capped, or hybrid arrangements such as success fees or retained advisory. AI’s contribution here is to make your team more efficient within that structure rather than to change the billing model. For SME clients with ongoing legal needs, subscription-style bundles are a growing option, covering a defined number of standard matters per month at a fixed monthly fee. LegalVision reports that its AI-supported chatbot resolves around 50% of SME queries without lawyer intervention, underpinning exactly this kind of portfolio pricing model.
When does value-and-risk pricing not apply?
Value-and-risk pricing does not replace hourly billing for all work, and getting this wrong in either direction costs you money. Hourly rates remain the clearer choice when scope is genuinely open-ended, the counterparty is unpredictable, or regulatory and third-party dependencies make cost forecasting unreliable. The discipline is identifying which work sits in which category before you write the engagement letter, not after the matter has run long.
Several counterforces could slow the shift at your firm. Clients with sophisticated procurement functions may push back, demanding that AI efficiency shows up as a discount rather than a maintained margin. In practice areas where both partners and clients are comfortable with hourly billing, the model is unlikely to change quickly regardless of what AI makes possible. If the change meets significant internal resistance, a mixed approach where you trial value pricing on new clients or standardised work first tends to be less disruptive than a firm-wide switch.
The other risk is under-scoping a fixed fee for work whose complexity you have not fully assessed. AI makes work faster, but it cannot make a difficult counterparty more predictable or a regulatory process shorter. Firms that lock in a fixed fee on a matter that then runs long often face a write-off that erases the efficiency gain they were trying to protect.
Only around 33% of UK firms provide lawyers with structured training in financial performance or matter economics. If you are changing your pricing model, the starting point is making sure the people setting and managing fees understand the cost base they are now pricing from, including what AI actually costs to run per matter.
What related factors shape what you can safely price?
Changing your pricing model for AI-enabled work touches several areas of professional practice that go beyond fee structure. UK data protection law, the SRA’s conduct requirements, and emerging AI regulation all affect what you can safely price at a low fixed rate versus what carries regulatory and liability exposure the price must reflect.
The ICO has published detailed guidance on AI and data protection under UK GDPR. If you feed client personal data into cloud AI tools, you need lawful basis, transparency with clients, and data processing agreements with your vendors. AI-assisted low-cost offerings must include clear explanations of human review levels and limitations, but those explanations cannot override your basic duty of competence.
The SRA Code of Conduct requires services to be competent, supervised, and clearly explained to clients, including any limitations on technology-generated outputs. The SRA has also published ethics and technology guidance warning that firms using technology must not mislead clients about who or what is performing the work. Any firm redesigning its service tiers around AI should read both documents before finalising engagement letter wording.
The UK government’s 2023 AI regulation white paper and the EU AI Act, formally adopted in 2024, are relevant for firms with any EU client exposure. Neither creates an AI-specific regulator for legal services yet, but the direction of travel is towards more transparency and accountability for AI-assisted work. The NCSC advises any organisation using generative AI to verify that its platform does not use client data to train public models and that data processing agreements cover this.
A clear pricing strategy communicates confidence. Clients who understand how you use AI, what it changes about the price, and what professional safeguards you maintain are more likely to stay than those left to assume you are cutting corners to cut costs.
