Enterprise
Law Firms Struggle to Prove AI Cost Savings to Corporate Clients
General counsel face CFO pressure for evidence that AI tools reduce legal bills, but outside firms lack consistent measurement frameworks.
Key takeaways
- Law firms can document AI tool usage but struggle to prove the technology delivers measurable cost savings to corporate clients.
- Five possible AI outcomes range from faster response times to no net efficiency gains when verification work offsets initial time savings.
- Current tracking methods—checkbox timesheets, narrative entries, informal notes—fail to connect AI use to reduced client bills versus enhanced deliverables.
- Four proposed solutions include standardized coding systems, matter-specific case reports, client collaboration on measurement frameworks, and vendor partnerships for data analysis.
- Firms that demonstrate concrete AI value to general counsel gain competitive advantage, while those offering only general assurances risk client relationships.
When chief financial officers ask whether artificial intelligence is cutting legal costs, general counsel need concrete answers. But law firms adopting AI tools are struggling to provide evidence that their technology investments translate into lower client bills.
According to James Dixon, a former chief operating officer at four Am Law 100 firms and current principal at Fidelium Advisors, most firms can document AI usage but cannot establish actual savings. That gap threatens client relationships as corporate legal departments face mounting pressure to justify outside counsel spending.
Why it matters
The disconnect between AI adoption and measurable client value creates a credibility problem for law firms at a time when corporate legal departments are under intense scrutiny to control costs. Firms that cannot demonstrate return on investment risk losing competitive ground to those that develop rigorous measurement systems—or face client skepticism about whether AI efficiency gains benefit the firm's margins rather than reducing bills.
Five Possible AI Outcomes
Dixon identifies five distinct results when law firms deploy AI tools: faster response times, reduced hours through more efficient workflows, expanded review scope examining full document sets rather than samples, novel insights from analyzing larger information volumes, or no net time savings when verification work offsets initial efficiency gains.
The challenge lies in connecting these outcomes to what clients actually receive and pay. A reduction in hours doesn't automatically lower bills when fee arrangements and scope of work remain unchanged. Broader document review at the same price may deliver value, but differs fundamentally from cost reduction.
Current Reporting Falls Short
Firms are taking fragmented approaches to tracking AI use. Some ask timekeepers to check boxes on timesheets indicating AI was used. Others request narrative descriptions of AI applications in time entries. Some encourage informal note-keeping about AI usage on individual matters.
None of these methods adequately measure whether efficiency gains survive the verification and correction process, or whether clients see reduced bills versus enhanced deliverables at unchanged prices.
Dixon notes that firm leaders frequently report clients value better outcomes over efficiency—a reasonable preference when stakes are high, but one that doesn't relieve pressure on general counsel to control costs.
Four Paths Forward
Dixon proposes four approaches for firms seeking to demonstrate AI value:
Standardized coding systems similar to the Uniform Task-Based Management System introduced in the 1990s could enable comparative analysis, though lawyer resistance to additional administrative burden remains an obstacle. AI itself might retroactively code time entries, but only if timekeepers consistently enter usage narratives.
Matter-specific case reports prepared at project completion or intervals would describe how AI was applied and identify outcomes, distinguishing measured savings from estimates and explaining what each benefit meant for the client.
Client collaboration invites corporate legal teams into the measurement process, allowing firms to understand what clients value most and demonstrate earnest effort. Avi Gesser, partner and co-chair of Debevoise's Data Strategy & Security Group, frames the challenge differently: "We're using AI to deliver a better service. It's like asking how much you saved by reading the NY Times digitally, rather than the physical paper."
Vendor partnerships with time and billing system providers or specialist companies could help capture and evaluate AI-usage data, comparing total hours on similar work including verification time.
The Competitive Stakes
Firms that can show what changed on a matter, what the client gained, and how they reached that conclusion give general counsel credible answers for CFOs. That evidence also supports pitches for future assignments.
Without demonstrated returns on legal spending, clients may question whether AI improvements benefit firm economics rather than their own. Firms that collaborate with clients on measurement frameworks can gain competitive advantage while laggards appear less capable or interested.
These findings were first reported by Bloomberg Law in an article by James Dixon.