The assumption driving most conversations about AI and legal pricing is straightforward: AI makes legal work cheaper to produce, so legal work should cost less. At the Financial Times Innovative Lawyers Global Summit yesterday, a panel featuring Priori CEO and co-founder Basha Rubin challenged that assumption from nearly every angle.
The honest answer, the session suggested, is that nobody really knows whether the total cost of legal services will go up or down, and the industry may be making significant commitments based on a premise that has not yet been tested.
The session brought together leaders from Prosus, the Association of Corporate Counsel, and RSGI for a wide-ranging conversation held under Chatham House rules. What follows reflects the substance of that discussion, including a notably engaged audience, without attributing specific remarks to specific individuals.
The opening frame: a picture more complicated than it looks
The conditions for pricing reform have arguably never been better: AI is reshaping how legal work gets done, in-house teams are building real capability, and the economic pressure on outside counsel spend has never been more visible. Yet AFAs still represent only five to fifteen percent of outside legal spend at Fortune 500 companies, even after fifteen years of sustained industry conversation. The AmLaw 100 has more than doubled revenue and profits per partner on rising hourly rates over the same period, and a recent study found that only four percent of in-house counsel report seeing any change in law firm pricing at all.
The panel’s diagnosis was not primarily about willingness. It was about structure, and about a set of deeply embedded habits on both sides of the relationship that no technology, on its own, will simply dissolve.
Will AI make legal services cheaper? Not necessarily.
The debate around legal pricing often starts with a simple assumption: AI reduces the hours required to produce legal work, and if firms pass those savings on, clients pay less. The panel examined that logic carefully and found several reasons to be skeptical.
First, firms have a credible counter-argument. High-stakes, high-judgment work may command premium pricing regardless of how efficiently it is produced. Law firm leaders the panel had spoken with believe they can charge more for existential matters, not less, precisely because clients value the outcome so highly. The car analogy offered in the room put it plainly: a Ford and a Bugatti may require similar manufacturing effort, but the Bugatti commands a fundamentally different price because of what it represents to the buyer.
Second, AI consumption costs are not fixed. As firms build AI into their workflows, they will face decisions about how to absorb and allocate those costs internally. The assumption that AI will always be cheaper than human labor in the long run is, as the panel noted, just that: an assumption. Pricing models for new technology always evolve, and consumption-based pricing for AI tools is coming. What that means for billing structures is genuinely unclear.
Third, and perhaps most counterintuitively, the total volume of legal work may be going up. More on that below.
The complexity problem no one expected
One of the session’s more provocative threads was the argument that AI may actually be adding to legal complexity rather than reducing it. Two macro forces, the accelerating regulatory environment and the proliferation of generative AI itself, are producing more legal work, not less.
Generative AI, the panel noted, is in many ways antithetical to standardization. Ask the same tool the same question three times and you get three different answers. That variability is finding its way into contracts and into litigation, raising new questions rather than closing old ones. Meanwhile, the expected explosion of pro se litigation, as AI lowers the barriers to filing lawsuits, was raised as a serious systemic concern. Every organization, the panel suggested, should expect to face more litigation in the coming years, much of it frivolous, and the court system is not yet equipped to absorb that volume.
The audience pushed back productively here, with several participants questioning whether the efficiency gains in high-volume, lower-complexity work would be enough to offset these pressures. The honest answer from the panel: probably not in the short term, and the firms and legal departments that assume a simple “AI in, costs down” dynamic are likely to be surprised.
Speed as a pricing dimension
A concrete example from the session reframed how several audience members thought about the value question. A firm used AI on a matter and saved its client a significant sum in fees, but the more consequential saving was five months of time. Those five months allowed the client to enter a market ahead of competitors. No fee reduction, however substantial, would have been worth the same amount.
The panel drew a distinction that resonated in the room: the pricing conversation in that case happened after the matter was complete, based on value actually delivered. That is a fundamentally different model from how most legal engagements are structured, and it requires a level of mutual trust that takes years to build.
The trust deficit, and what to do about it
This became one of the session’s central themes, and audience members returned to it repeatedly. A story of an in-house counsel describing the pressure she was under to use a specific AI tool for work it genuinely could not yet handle.
That example opened a broader conversation about governance. The panel was direct: governance structures are not yet in place at most firms, and the gap between what AI is being marketed to do and what it can actually deliver reliably is creating real risk. High-profile hallucination incidents, including a case in which a brief filed in New York contained dozens of fabricated case citations, were cited as the kinds of mistakes that tend to accelerate institutional response.
The prediction from the panel: more governance frameworks will emerge over the next twelve months, not because the industry is proactive, but because the cost of not having them is becoming impossible to ignore.
From the floor, one participant put it plainly: the industry has been too unbridled in its expectations of what AI can do today.
RFPs, transparency, and the budget disclosure problem
An exchange from the audience surfaced a structural inefficiency that many in the room recognized immediately. In-house teams routinely withhold budget information from RFPs, a practice that, while understandable tactically, produces fundamentally misaligned responses. Firms quote without knowing whether they are competing for a Mini or a Bentley, to borrow the analogy offered in the room.
The panel’s view: stating a budget ceiling upfront is increasingly common among the most sophisticated legal teams, and the data to support better upfront pricing is becoming available on both sides. AI-driven analysis of historical billing patterns can help firms price more accurately at the outset, rather than discovering misalignment after significant time has been invested by everyone.
Value-based pricing requires a different kind of relationship
The session closed by returning to the question it opened with. AFAs are a structural adjustment. Value-based pricing is a different thing entirely; it requires untethering the fee from the input and anchoring it to what the outcome was actually worth to the client. That is a harder conversation, and it only happens in relationships where both parties trust each other enough to be honest before, during, and after a matter.
The panel’s collective view was that the legal industry has the tools, the data, and in many cases the inclination to get there. What it still lacks, in most client-firm relationships, is the habit. Building that habit was the challenge the audience left with, and based on the energy in the room, one that more people are taking seriously than the AFA adoption statistics alone might suggest.
Learn More
- See how Priori RFP helps legal teams make more informed outside counsel decisions with AI-powered RFP workflows, pricing insights, and historical benchmarking.
- Download The Pre-RFP Playbook to learn how AI-enabled intake and structured sourcing workflows create more consistent, transparent outside counsel decisions.
- Explore our perspective on incentive alignment and legal pricing to understand why pricing models alone cannot solve deeper structural challenges between clients and law firms.
- Join our next Advisory Office Hours as we continue the conversation on legal pricing, AI, and the future of outside counsel management.
- Continue the conversation at FlexFest on July 29, where legal leaders will explore how AI is changing legal pricing, work allocation, and the operating models behind the modern legal department.
