The billable hour model has been under client pressure for years. In-house legal departments want cost predictability. Procurement teams want to negotiate fixed prices, not open-ended hourly exposure. Corporate clients have been pushing for AFAs since the 2008 financial crisis, when outside counsel spend became a target for cost reduction.
Law firms have largely resisted, for a simple reason: fixed fees are financially dangerous when you cannot accurately predict how long a matter will take. If a firm offers a $5,000 fixed fee for a commercial contract negotiation and the matter consumes 30 attorney hours, the effective rate is $167 per hour — far below what most firms need to operate profitably. The risk of scope creep, client intransigence, and complexity underestimation has made law firms cautious about committing to fixed prices.
AI changes this calculus in a meaningful way. For categories of work that AI can assist with consistently — NDA review, standard commercial agreements, routine employment matters, immigration filings — AI reduces both the time required and the variability in that time. The range of possible outcomes narrows. Fixed pricing becomes predictable rather than speculative.
This is accelerating AFA adoption in ways that client pressure alone never achieved. When AI can complete a first-pass NDA review in 30 minutes with consistent quality, a firm can confidently offer a $300 fixed fee, cover its costs, and still deliver value to the client — without the financial risk that made fixed fees unattractive for the past three decades.
How It Works
Types of AFAs and Their AI Fit
Alternative fee arrangements span a range of structures, each with different AI applicability:
Fixed fees set a single price for a defined scope of work. AI makes fixed fees viable by compressing completion time and reducing variance. Best applied to high-volume, well-defined tasks.
Capped fees use hourly billing up to a maximum. AI helps firms set realistic caps by enabling accurate bottom-up cost estimation for each component of the matter.
Blended rates charge a single hourly rate across all timekeepers, regardless of seniority. AI can reduce the need for junior attorney hours, which shifts the effective blended rate calculation.
Contingency fees remain fee value-based. AI's primary impact here is on case evaluation: AI litigation analytics tools can inform settlement value assessment and case selection.
Retainers charge a monthly or annual fee for defined services. AI enables firms to fulfill more retainer work with fewer attorney hours, improving retainer profitability — or to offer lower-cost retainers competitively.
The Cost Prediction Improvement
AI improves AFA pricing primarily through two mechanisms. First, AI-assisted work completion is faster and more predictable than manual work, giving firms better data on actual cost per matter type. A firm that runs 200 NDAs through AI review over six months has reliable data on actual AI-assisted completion time, which enables confident fixed-fee pricing.
Second, AI analytics within practice management platforms like Clio and enterprise tools like Onit enable historical matter cost analysis — pulling data on what comparable matters actually cost, which provides a basis for fixed-fee calculations.
Client Negotiation Dynamics
Large clients negotiating AFAs now frequently ask whether and how AI is being used in their matters. A firm that uses AI to complete work faster and then prices a fixed fee based on pre-AI costs is capturing a margin the client views as excessive. Client sophistication about AI is growing; AFA pricing negotiations increasingly include explicit discussion of AI's role in task completion and cost structure.
Filevine and similar platforms provide matter analytics that give firms the data needed to negotiate AFAs from evidence rather than intuition.
Key Considerations for Law Firms
Scope definition is everything. Fixed fees are only profitable if scope is precisely defined. AI efficiency gains are irrelevant if a fixed-fee matter scope expands beyond the original definition. Engagement letters for fixed-fee matters must define scope precisely, including what happens when scope expands: does the fixed fee increase? Does billing revert to hourly? Clear scope discipline protects the economics of AI-enabled fixed pricing.
Track actual AI-assisted time. Firms implementing fixed fees based on AI efficiency assumptions must verify those assumptions with time data. If the assumption is that AI reduces NDA review from 3 hours to 45 minutes, track actual completion times across the first 50 matters and compare to assumptions. If assumptions are wrong, reprice.
Do not underinvest in AI infrastructure. Fixed-fee economics depend on AI actually performing as assumed. AI tool downtime, version changes that reduce accuracy, or inadequate attorney training can all cause actual completion times to exceed fixed-fee assumptions. The AI infrastructure that underpins AFA pricing is a business-critical dependency.
Consider matter type carefully. Not all matter types are AFA-ready. Litigation, regulatory matters, and complex transactions involve too much uncertainty for most fixed-fee structures. Firms should identify the specific matter types where AI enables reliable cost prediction and target AFA offerings at those categories first.
Address AI costs in the fee structure. AI tool costs — licensing, per-use fees, and implementation — are real costs that need to be recovered in AFA pricing. Firms must decide whether AI costs are embedded in the fixed fee or billed as a pass-through expense. Client expectations and engagement letter language should be consistent on this point.
Limitations and Risks
Scope creep erodes fixed-fee profitability. The most common failure mode for fixed-fee arrangements is scope expansion that is not addressed by fee adjustment. AI efficiency gains are absorbed by scope creep, and the firm ends up at or below break-even. Scope management discipline is a prerequisite for successful AFA programs.
AI performance variability. AI tools do not perform identically across all documents. A standard NDA from a repeat counterparty may take 30 minutes to review; a heavily negotiated, unusual NDA from a first-time counterparty may take 2 hours even with AI assistance. Fixed-fee pricing based on average AI performance is still exposed to outlier cases.
Competitive undercutting. If AI makes legal work genuinely cheaper, competitors may offer fixed fees at prices below current firm costs. Firms that delay AI investment may face competitive pressure from AI-enabled competitors offering significantly lower fixed fees for the same work.
Client resistance to AFA limitations. Clients may resist scope limitations that protect the firm's fixed-fee economics. A client who views a fixed-fee engagement as covering unlimited revisions will conflict with a firm that has budgeted for two rounds of revisions. Clear engagement letter drafting and proactive scope management are required.
Data quality for pricing. AFA pricing based on historical cost data is only as reliable as the underlying data. Firms with incomplete time tracking — a common problem — may set fixed fees based on cost data that is 20-30% understated, leading to mispriced agreements.