Is Your Firm's AI Adoption Quietly Costing You Revenue?

Law firms are moving fast on generative AI. And most are budgeting for it like any other software line item. That's the problem.

Legal AI spend differs from other tech costs because it directly conflicts with the billable hour. When an associate uses AI to draft in 45 minutes what used to take three hours, the firm doesn't just save time: it can lose the revenue that hourly billing was built to capture, unless pricing has been redesigned around it.

At the same time, vendor pricing is opaque and consolidating fast. Harvey, the best-funded legal AI vendor, still won't publish its pricing. Bundling deals like its LexisNexis content partnership are expected to push all-in costs materially higher for firms that want the combined offering: often after the contract is already signed. Layer a frontier-model token bill on top of an existing legal tech stack, and total technology spend can balloon well past what was modelled at the start of the year.

For Australian firms, there's a third layer. Regulators, including the NSW Law Society, the Victorian Legal Services Board and Commissioner, and the WA Legal Practice Board, require that billing reflect the actual legal work performed, even where AI is used. An uncontrolled AI cost blowout passed on to a client isn't just a pricing dispute; it risks a conduct complaint.

None of this is an argument against adopting AI. It's an argument for governing AI spend with the same discipline firms already apply to matter budgets and realisation rates before the invoice arrives, not after.

Receive Our Report and Checklist Guide

We've put together a practical governance guide and checklist covering what to lock down before rollout, how to set access and usage controls, how to monitor spend, how to realign billing to capture AI's value instead of losing it, and how to keep vendor contracts under control at renewal.

Subscribe to receive the complete report and guide in our upcoming newsletter.