It took me the better part of last weekend to process everything I'd seen and heard at ILTACON. No, I'm not talking about the booth from one legal AI vendor that was larger and furnished much better than my first apartment. It has more to do with where legal technology is right now, in relation to firms and what they are trying to do with it, and where it needs to go next to deliver the kind of value firms will need to remain competitive and profitable.
Not surprisingly, all these thoughts lead back to data. The education sessions at ILTACON, which are top-notch because they are peer-created and peer-led, explored different facets of data in the AI-enabled firm. Information governance sessions provided practical solutions for managing data to policy for retention, storage, and compliance. The information technology sessions focused on keeping it secure through a combination of vendor management, data governance, and security approaches. The knowledge management topics went deep on how to organize and classify information for AI access with taxonomies and structures. Several data science discussions dove into the emerging role of legal data scientists and their work preparing and querying data for analysis.
Two things stand out. The first is that these sessions were in separate rooms, drawing attendees aligned with the role of the speakers. The second is what most of them left out: the client. The discussions centered on the strategic and tactical work of managing data in an increasingly AI-driven context. The conversations focused on the work of the firm: keeping the data perimeter secure, getting timekeepers to use it to make their workflows more efficient, controlling the access and flow of matter data, making it easier for AI to process, ensuring its compliance with policy and client guidelines, and leaving a defensible audit trail. What I didn't hear anyone talk about is using data to deliver what clients want.
Clients hire firms for their experience and deep expertise, honed from working across the firm's collective clients and their matters. They seek legal strategy, expertly applied, to achieve specific outcomes. They are buying firm-level intelligence, and more clients are questioning how AI can and should be used to deliver it. Understanding that AI can make tasks and processes faster and more streamlined, clients are increasingly expecting firms to use it to deliver the work as efficiently as possible and for their bills to reflect it. This piece is operational- and practice-level efficiency, driven by AI, where clients are looking for fewer billable hours and faster delivery.
But there's a catch. Although clients want to tap into the firm's work across all clients and matters, they don't want their data contributing to that AI-enabled knowledge base going forward. Outside counsel agreements increasingly specify how and where a client's data is stored and processed, with many stipulating it may not be used for training AI models, let alone accessed by other attorneys or used on other clients' work. This position puts firms at a distinct disadvantage from corporate legal departments, which hold data for the organization as the singular client. Free from the constraints of ethical walls and conflicts, corporate legal teams can now collect, mine, and model their data to produce meaningful outputs that drive insight into legal strategy, workflow improvements, and case management approaches. In contrast, within the client confidentiality and ethical wall constraints, a firm can only point AI at the siloed slices of data their practice groups are allowed to reach.
Prior to AI, the collective firm-level knowledge and how to apply it lived in partners’ heads. The shift to generative AI is placing that knowledge, in the form of client data, into the DMS, a curated AI library of approved content in a staging area, the knowledge management platform, and at some firms a data lake house blending structured and unstructured data. As we heard in the sessions at ILTACON — retention, privacy, ethical walls, access control, and audit trails to show compliance — every one of those controls comes back to matter.
Firms manage data by matter. Whether a file can be deleted, leaves with a departing partner, or sits behind an ethical wall comes down to which client matter it belongs to and whether that matter is closed. The tools filling the exhibit hall answer other questions at the operational or practice level: who touched a file and when, how sensitive it is, how it relates to an earlier agreement or redlined version, the relevance of the research underpinning the work, and where the information is stored securely for permissioned access. Those questions matter to IT and the CISO, the timekeepers producing the work, and the information governance and knowledge management teams who govern and organize the data. The existing tools in the legal tech stack answer these questions well and are improving the operational and practice area work of the firms investing in them. All of these solutions offer a window into the data at a matter-specific level. None of them independently will deliver the firm-level insights clients are seeking and that AI is increasingly able to provide. It’s both the next-level AI challenge and a significant opportunity for firms to deepen their value proposition as corporate legal departments increase their capabilities.
The question I'd like to see ILTA members explore at ILTACON 2027 is how firms can use AI to deliver what clients want: the entirety of the firm's knowledge and experience across clients and matters, delivered within the clients’ data constraints. Is it a knowledge layer populated by an analysis of the raw client data that contains no distinguishable client information? Perhaps. Are there other approaches worth debating? Plenty. I'm not going to pretend to know the answer, because it will take stakeholders from across information governance, compliance, information technology and security, knowledge management, and data science to figure it out. Bringing those stakeholders together is the point of ILTACON.
What I do know is that understanding what data you have, where it lives, and what matter it belongs to is foundational to the solution, whatever form it takes. My hunch, based on the conversations I had with CIOs, data scientists, knowledge management, and information governance leads is that the dark data repositories, the file shares, OneDrive, Exchange, and DMS personal workspaces contain valuable information. The kind of firm-level information that lived in people's heads when the AmLaw 100 firms built the experience and expertise that grew them into the industry's leaders. The next evolution for these firms is to capture that knowledge and use it in a way corporate legal departments can't, against similar work spanning thousands of clients and hundreds of thousands of matters. It's a challenging problem and solving it will carry AmLaw's application of AI beyond doing legal research, drafting documents, and comparing contracts into its full potential of reshaping the future of legal work. That conversation is the one I'm looking forward to having at ILTACON 2027, with everyone in the same room and the client at the center of it.