Legalweek 2026 Session Recap

Key Takeaway: Compared to other technologies, AI requires a more nuanced approach to questions of the Total Cost of Ownership (TCO) and ROI. Panelists at the opening session of Legalweek 2026 discussed why deploying AI incurs greater costs than other tools, and the key consideration to achieve favorable ROI for AI as quickly as possible.

As AI and advanced technologies reshape legal operations, Total Cost of Ownership (TCO) has evolved from a static cost model to a dynamic value framework. Legal departments and law firms are wrestling with how best to assess hidden costs, ROI, and strategic impact to make smarter technology investments. This was the focus of the Legalweek 2026 legal operations workshop session presented by Epiq, Measuring Total Cost of Ownership in AI-Driven Legal Operations. Jon Lavinder, Senior Director of Product Management at Epiq, moderated a panel featuring:

• Sandra Metallo-Barragan, eDiscovery Counsel, Proskauer 
• Clinton Sanko, Practice Enhancement and eDiscovery Officer, Baker Donelson 
• Amy Sellars, Of Counsel, Gunster
• An Trotter, Senior Director of Operations, Office of General Counsel, Hearst

Requisite Shifts in Thinking for AI Adoption

Jon began the session by noting that while cost calculation for other IT tools is straightforward, AI involves numerous “below the waterline” costs beyond acquisition, implementation, operational, and training costs. With AI, additional costs arise from governance, policing shadow AI, piloting, allowance for risk and error, and flexibility as a hedge against provider lock-in. Moreover, the training costs for AI vastly exceed those for other technologies. 

Expanding on the training costs of AI, multiple panelists emphasized that effective use of AI requires, above all, instilling a new mindset among lawyers. As Amy put it, “learning to use the tool is not as important as changing the way one thinks about how tools will evolve. It’s crucial to teach people how to think about AI.” An echoed that point, highlighting the importance of learning how AI works at a conceptual level. This drives additional training costs, because becoming proficient with AI requires an iterative approach to build competence with it. 

Piloting AI is crucial. An recounted that in-house, hands-on “agent builder days” are invaluable in prompting the “aha” moments that reveal to users the power and possibilities of AI. 

Adequate AI training is not a one-off. As Amy explained, the only way to identify where AI can add value is by tinkering with the tools, and lawyers must receive the time and latitude to engage in cycles of experimentation, failure, and adjustment that iteration requires. Sandra stressed that this cannot be rushed or forced. The initial experiences of AI users are a decisive influence on their likelihood of embracing the tool.

The Challenges of Measuring Total Cost of Ownership and Return on Investment

In discussing ROI, the panelists differentiated between costs (i.e., investment) and value (i.e., return). For AI to achieve favorable ROI, firms must reduce ownership costs and focus on use cases that deliver the greatest value.

Clinton remarked that a crucial step in reducing AI TCO is determining use cases that make the most impact on the greatest number of lawyers. For different practice groups, that will vary. To deliver maximum value, it is important to address use cases that existing tools cannot. 

However, measuring ROI will remain a challenge. Jon noted that with use cases such as eDiscovery, AI’s value-add is relatively easy to quantify. That said, as Sandra explained, assessments of AI’s time-to-value for other types of work remain largely anecdotal. Clinton stressed that analyzing AI ROI is best done across a portfolio of similar matters. This provides greater insight into the use cases that deliver the greatest benefit to the most lawyers at the lowest cost while balancing risk and the drive to deliver value to clients as rapidly as possible.

Beyond the Total Cost of Ownership

There was agreement amongst the panelists that, strict measures of TCO and ROI aside, there are strategic business considerations that cannot be measured. With law schools now teaching students how to integrate AI tools into the practice of law, young lawyers now expect firms to not only provide AI tools but also have an innovative and impactful use of AI. At the same time, this AI-savvy cohort adds immediate value to more senior lawyers by assisting them in adopting AI. That makes deploying AI a matter of competitive necessity in attracting and retaining top legal talent. 

At the same time, client expectations are shifting. With, by one count, nearly 200 AI tools released in 2025, the pressure to quickly deploy the technology has become intense. Clients not only expect firms to use AI but also to deliver an unprecedented increase in responsiveness. Sandra cautioned that in the face of such rising expectations, firm lawyers may need to level-set expectations around speed of response and AI’s capabilities. 

TCO and ROI for AI will remain a complex matter. Data governance, appropriate safeguards, the rapid evolution of AI tools, and improvements in the ability to measure ROI for specific use cases will all impact both ownership costs and ROI. The value of AI for legal is now beyond dispute. The question is no longer whether to do so, but how.