By Gravity Stack Staff
Across the legal industry, AI has moved from experimentation to line item. Model companies have raised unprecedented capital. Law firms and legal departments have begun allocating meaningful budget to pilots, licensing, workflow redesign, and internal enablement. With this level of investment, the question that inevitably surfaces is some version of what is the ROI?
Recently, I joined a panel at Legal Innovators New York titled “Legal AI Is For Real – What Now?” where I shared insights with panelists Greg Baker – Global Head of Practice Innovation, Linklaters, Shruti Ajitsaria – Partner and Head of Fuse, A&O Shearman, Josh Rosenzweig – Senior Director of AI and Innovation, Morgan, Lewis & Bockius LLP, and Max Junestrand – Co-Founder & CEO, Legora on just this topic.
But going deeper than you can on a panel, I wanted to jump back to our opening thought…It is a fair question, but it often presumes a clarity that does not yet exist. Traditional ROI frameworks were built around systems with predictable inputs, stable workflows, and measurable efficiency gains. AI does not conform to that architecture. It behaves more like a capability layer than a discrete tool, and early attempts to measure it have struggled to account for its distributed impact across knowledge work.
The MIT finding that ninety five percent of enterprise AI pilots fail is frequently cited as evidence that the technology is oversold. A more accurate interpretation is that most pilots were structured around evaluation methods inherited from earlier generations of enterprise systems. They search for linear causality in environments where the benefits are rarely linear. They focus on task time reduction when the real value appears in decision quality, reduced rework, accelerated throughput, or entirely new analytical avenues. The limitations lie not in the technology but in the frameworks used to assess it.
This is also why familiar analogies persist. Asking for the ROI of AI often resembles asking for the ROI of a second monitor. Every attorney and knowledge worker understands the improvement in focus, fluidity, and cognitive load. The impact is immediate, meaningful, and consistently felt, but it is rarely captured cleanly in a financial model. The benefit is real, yet distributed across thousands of micro-decisions that enable better work.
AI has always followed a consumer-first adoption pattern. The return is employee first and enterprise second. In legal, this takes shape in how quickly people can synthesize context, explore alternatives, or evaluate risk with greater consistency. The uplift does not sit neatly on a balance sheet, yet it reshapes how work gets done. The most forward-leaning firms and legal departments recognize this and focus on equipping their people with the best available tools rather than waiting for an all-encompassing ROI calculus to mature.
Next year, we will begin to see more sophisticated approaches to P&L-level measurement. As processes become more agentic, workflows more automated, and throughput more measurable, financial signals will become clearer. Margin effects, cycle time compression, and revenue enablement will start to register. But that phase is downstream from where most of the market is today.
At this moment, the organizations making the most progress are those treating AI as part of the operating environment rather than as an isolated point solution. They understand that the real competitive advantage lies in enabling people to work at a higher level of judgment, speed, and creativity. They measure the outcome of that empowerment, not just the mechanics of a single task.
At Gravity Stack, we continue to test, evaluate, and operationalize emerging AI capabilities so our clients do not have to navigate this transition alone. If your organization is reconsidering how it evaluates AI impact, or how to align measurement with the realities of knowledge work, our team is here to help.
👉 If your department is exploring AI, get in touch.
—
Follow Gravity Stack on LinkedIn for updates from our AI Lab, client stories, and legal innovation insights.
