Reports surfaced this week that Meta has told outside counsel it will not pay law firm rates for work that AI can already do. Zscaler and UBS appear to be drifting toward the same position. None of these companies has published a policy framework. They probably don’t need one. The message is simple enough to fit in a renewal email: if the task is automatable, the firm doesn’t get to bill it like it isn’t. 

It would be tempting to read this as a single client getting tough on price – move on, nothing to see here. It’s more accurate and more useful to read it as the billable hour’s most reliable customers finally pricing in what they already suspected. Corporate legal departments have spent two years watching AI compress drafting, research, and document review into a fraction of the time those tasks used to take. They didn’t need a vendor to explain the math. They just needed the data to prove what they suspected and the leverage to say it out loud. Meta found the leverage first. 

A recent Legaltech Hub piece on this exact tension makes the more important point. The obstacle isn’t that firms don’t want to change. Most are deploying AI, running pilots, and genuinely trying. The obstacle is that the entire law firm model — compensation, evaluation, pricing, even how partners get credit for what they build — is calibrated around time as the proxy of value. Take time away as the unit of value, and nothing else in the system knows what to do. Take the old management axiom – you get the behaviours you reward. Let’s say an associate builds a workflow that cuts a recurring task by eighty percent. They haven’t billed more hours; quite the reverse – they have eliminated the opportunity for hours. In most firms’ compensation models, that contribution is functionally invisible. The firm wants efficiency in order to compete. Yet the firm has no mechanism to reward the person who created the efficiency gains.  

So that’s the rock. Meta is the hard place. A client telling you not to bill for AI-assisted work isn’t asking you to discount your rates. It’s asking you to reprice value in a system that was never built to measure value independently of hours. Firms can’t just say yes to that request, even if they genuinely want to. Saying yes requires answering questions most firms haven’t touched: what is this work actually worth if not its duration, how do we compensate the people who build the systems instead of the people who log the hours, and how do we explain any of this to a partner who has benefitted so handsomely under the old mechanisms. 

So here we come back to AI – the tool selection was never the hard part. Pricing model design, role redefinition, and the governance to back both up are the hard parts, and they’re squarely structural and operational problems, not technology problems. Solving them piecemeal, one matter or one client demand at a time, just produces a patchwork of exceptions that nobody can defend in front of the next general counsel who asks the same question Meta just asked. 

The firms that get ahead of this won’t be the ones with the most AI licenses. They’ll be the ones who did the unglamorous work first: figuring out where AI-assisted delivery actually creates measurable value, building a pricing structure that can hold up when a client pushes on it, and giving people outside the billable-hour track — the associate who built the workflow, the non-lawyer who maintains it — a real stake in the outcome. None of that is a pilot project. It’s a redesign, and redesigns go better with someone outside the partnership doing the diagnostic work, because nobody inside a billing structure is well positioned to mark their own homework. 

This is the work we do at BLS Consulting. Not implementing AI tools for their own sake, but helping firms work out what their AI-enabled delivery is actually worth, where that value sits outside the traditional matter and billable hour, and how to build pricing, roles, and governance that can survive a client like Meta asking the question directly. Technology is the means. The point is removing the waste — billing waste, training waste, the lawyer-hours spent on what a system could absorb — that keeps good lawyers from doing the work only they can do, and keeps clients from getting it at a price that makes sense. 

Meta isn’t going to wait for the legal industry to finish debating whether the billable hour is dying. Neither will the next client who reads the same numbers and reaches the same conclusion. The firms working through that question now, deliberately and structurally, will be answering it on their own terms. The ones waiting for clarity will be answering it on Meta’s. 

Andrew Terrett – BLS Consulting — Thought Leadership (July 2026)