The solicitor who reduced her standard commercial lease review from four hours to forty minutes using AI gets one question from almost every new client: if the computer did most of it, why has the fee not changed? It is a reasonable question. The honest answer is that the fee was never really about time. It was about the value of getting the outcome right, and who carries the cost if it goes wrong.
AI has compressed delivery time across legal and advice-driven services, but the underlying economics of risk and client value have not moved with it. Firms pricing on hours alone are now sitting on a structural problem. Those that reframe their fees around value and risk are finding a more defensible position, and in many cases a more profitable one.
What is value-and-risk pricing for AI-enabled legal work?
Value-and-risk pricing anchors a fee to two things: the commercial value of the outcome to the client, and the regulatory and professional risk your firm is carrying in delivering it. A commercial property transaction for a business acquiring its fourth site carries more value and more exposure than a one-off NDA. Those factors do not shift when AI speeds up the drafting.
The move away from hourly billing in legal services predates AI. Alternative fee arrangements (AFAs) have been growing for years. Thomson Reuters describes fixed-fee arrangements and success-based pricing tied to business outcomes as the main alternatives for AI-enabled work. What AI does is make those models commercially safer to offer. When AI can process a routine contract review in a fraction of the time, a firm can quote a fixed fee without guessing at hours.
Value tiers work like this in practice: fees are set against the commercial significance of the matter rather than the time taken. A deal-size bracket for significant transactions, a per-document volume rate for due diligence, a success component for disputes resolved or penalties avoided. Risk multipliers sit on top: FCA enforcement exposure, special-category personal data, cross-border regulatory complexity, and whether the matter involves AI systems classified as high-risk under the EU AI Act.
Why does this matter for your firm?
The commercial pressure is already building. A 2024 Axiom survey found that while 79% of law firms now use AI, 45% of in-house counsel believed outside firms were charging them more because of AI, and 68% planned to negotiate AI-related pricing in the next twelve months. If you cannot articulate what drives your fees, your clients will drive that conversation for you.
There is also a margin case. Thomson Reuters documents firms using AI to support fixed fees and outcome-based pricing, aligning efficiency gains with client value rather than discounting. A 2023 Fennemore analysis forecast that Alternative Fee Arrangements would grow from 20% of law firm revenue in 2023 to over 70% by 2025, driven by automation and client demand for cost certainty.
The risk exposure is equally concrete. Following the 2023 case in which two New York lawyers were sanctioned for submitting ChatGPT-generated case citations that turned out to be fictional, UK legal commentary has consistently flagged that human validation of AI output is not optional. Pricing that assumes zero review time creates a malpractice exposure and, where client data is involved, a data-protection liability under the ICO’s framework. Your fees need to cover the oversight layer that AI-enabled delivery makes necessary, not just the production time.
Where will you actually meet this in the UK market?
The pattern is visible at the top of the market and it filters down. Linklaters launched its MatterExplorer AI platform in 2024 to support fixed-fee alternatives for large transactions, positioning AI as what makes predictable pricing commercially safe. Allen & Overy, now A&O Shearman, has publicly linked its Harvey AI deployment to client conversations about fixed-fee and outcome-based work. The underlying logic applies to a twelve-person practice as readily as to a Magic Circle firm.
Luminance, the UK-founded AI contract analysis provider, prices on document volumes and subscriptions. That model lets firms convert variable review effort into predictable packages, which makes fixed-fee quotes to clients reliable rather than speculative. The tools and the pricing logic are not restricted to large firms.
The regulatory context reinforces the risk-based framing. The ICO’s guidance on AI and data protection states that organisations remain responsible for processing outcomes when using third-party AI tools, and that liability cannot be outsourced to the vendor. The FCA’s Consumer Duty, in force since July 2023, requires that clients receive fair value and that firms prevent foreseeable harm. Where AI is used on personal data, particularly special-category data or anything connected to FCA-regulated products, the risk does not reduce because a machine performed the first pass. In 2022 the ICO fined Tuckers Solicitors LLP £98,000 following a ransomware attack that exposed court bundles, citing failures to encrypt personal data. AI tooling that centralises large volumes of sensitive client data can increase the potential impact of such incidents considerably. Pricing higher-risk matters accordingly gives you the margin to fund the controls those matters require.
When does it work, and when does hourly still make sense?
Value-and-risk pricing works when scope and risk can be defined before the matter starts. Routine commercial contracts, employment agreements, standard data-protection documentation, and matters with clear precedent are good candidates for fixed or capped fees. Matters where scope is genuinely unpredictable, where litigation discovery changes week to week, or where client instructions keep shifting are harder to price in advance and often still warrant time-recording as the underlying discipline.
The practical division: use fixed fees for high-volume, lower-risk work where AI delivers large efficiency gains and scope is clear. Use value-banded fixed fees with deal-size brackets for significant transactions. Use time-and-materials with a cap for genuinely open-ended matters. Apply risk multipliers across all three categories: FCA enforcement exposure, EU AI Act high-risk classification, cross-border data transfers, and elevated cyber-risk profiles. The ICO and FCA guidance gives you the language to explain those multipliers to clients. Reference it in your engagement letters.
The client-resistance counterpoint is worth naming. Axiom’s data shows 68% of in-house counsel planning to negotiate AI-related pricing. That pressure is real. The answer is not to absorb it by discounting quietly. It is to be able to explain precisely what the fee covers, what risk your firm is carrying, and what the oversight and governance layer actually costs to run properly. Clients who understand what they are paying for are more straightforward to hold a position with.
What does the practical shift look like?
The starting point is simpler than a full pricing redesign. You are adding two questions to how you quote a new matter: what is this outcome worth to the client, and what risk is our firm taking on? For the matters you handle most frequently, you can answer both in ten minutes at intake. That is enough to stop pricing on the clock alone.
Group your common matter types into three tiers: standard work at fixed fees (NDAs, employment contracts, routine data-protection documents), significant matters at value-banded fixed fees, and complex open-ended matters with time records and a cap. Then build a short risk-multiplier checklist for intake: does this matter involve special-category personal data, FCA-regulated products, EU AI Act high-risk AI systems, or elevated cyber exposure? Where the answer is yes, the fee goes up by a defined uplift and the reason goes into the engagement letter.
Price human review of AI outputs into the base from the start. A standard review step or named-partner sign-off is not optional overhead. Thomson Reuters notes that reliable AI-enabled legal work requires human expertise at the review stage. Clients whose expectations do not reflect that need to understand it before the engagement starts, not after the invoice lands.
If you want to work through your pricing structure alongside reducing your own dependency in the practice, that is the kind of work the Founder Freedom Programme is designed for. Book a conversation.
