At Mitratech Interact, a Holland & Knight partner and a Plains All American legal executive described the decisions about scope, cost and judgment that an instruction to use AI leaves unresolved.

An AI-assisted comparison of 40 precedent credit agreements took Parker Young 30 hours. In her judgment, its marginal benefit did not justify the cost. The client had expected the technology to make the exercise cheap; the client’s own customer saw no value in the result.

At Stewart Title, Sharon Morrison built support for a stalled legal systems replacement by connecting executives’ need for better information with employees’ knowledge of how the work gets done.

Stewart Title’s business leaders wanted answers its legal systems could not readily provide. They needed faster responses and a clearer view of claims losses coming into the next quarter. The CEO’s assessment, as Sharon Morrison recalled it, was direct: “I don’t know what’s going on in your department. I need this information.”

A four-person legal operations team expanded intake from three workflows to 14 by studying what employees submitted and where the original design fell short.

AI questions were turning up in Delta Dental’s RFP workflow, and some had nothing to do with an RFP. The workflow carried a supplemental request type for AI, the only AI request type in any of the division’s workflows. Because people knew where it was, that is where they went.

Eudia’s Omar Haroun says procurement and revenue teams are contributing increasingly to usage and growth, broadening the demands on technology developed for legal work.

The acquisition combines Repario’s discovery and infrastructure business with UnitedLex’s wider legal services operation. How the two will be integrated, and what changes for clients, is not yet spelled out.

Legal departments have spent years measuring law firms after the work is assigned. AI is bringing that accumulated performance record forward, to the point where the next assignment is made.

The knowledge was already there. What changed is the cost of finding and using it. At RLLB, leaders from Equinix, Cargill, GoPro and Phillips 66 showed how decades of contracts, disputes and risk decisions can become an advantage for the business.

Legal has learned to measure AI against the hours it saves. Agents complicate that math by performing work no legal team would have commissioned from humans in the first place.

At NEXT Insurance, iFood and Expedia, faster legal work depends on a continuing investment in the knowledge behind the answers. That investment is part of AI’s economics.