Introduction

Buying legal software has always been straightforward. If a firm has 100 lawyers, it buys 100 licences. The more people who use the software, the more the firm pays.

Legal AI is changing that. Unlike traditional software, AI costs money every time it completes a task. The more work you ask it to do, the more expensive it becomes to run. That means charging firms simply based on the number of users no longer makes much sense.

Why Doesn't Per-Seat Pricing Work?

Traditional software is relatively cheap to provide once it has been built. Adding another user usually costs the software company very little.

AI is different. Every prompt uses computing power, and some tasks require far more of it than others.

For example, asking AI to summarise a document is fairly inexpensive. Asking it to review hundreds of contracts or analyse thousands of pages of evidence is much more demanding. Two lawyers with the same subscription could therefore cost the AI provider completely different amounts to support.

This is why the number of users is becoming a poor way to measure how much a firm should pay.

So, How Might Firms Be Charged Instead?

Legal AI providers are already testing new pricing models that better reflect how the technology is used.

Some possibilities include:

  • Paying for the number of AI requests made.
  • Buying a pool of credits that everyone in the firm shares.
  • Paying a monthly subscription with extra charges if usage exceeds a set limit.
  • Charging for completed legal tasks, such as reviewing a contract, instead of access to the software itself.

Each approach tries to link the price more closely to the actual cost of running AI.

Why Does This Matter?

For law firms, paying per lawyer has always made budgeting easy because costs rise alongside headcount.

Usage-based pricing changes that. A quiet month with little AI use could cost very little, while a large transaction or litigation involving thousands of documents could generate significantly higher AI costs.

Instead of treating AI as another software licence, firms may begin tracking AI spending for each client matter.

How Legal Teams Get Involved

Changing how legal AI is priced raises new legal and commercial questions. Different teams within a law firm are likely to become involved.

Commercial Team

  • Negotiate pricing models, usage limits and payment terms.
  • Review contracts dealing with additional charges and service levels.

Competition Team

  • Consider whether a small number of AI providers could gain too much market power.
  • Advise on competition law issues if pricing practices become anti-competitive.

Regulatory Team

  • Monitor new AI regulations affecting legal technology.
  • Advise firms on compliance when AI becomes part of legal workflows.

Risk and Compliance Team

  • Develop policies for monitoring AI usage.
  • Ensure firms can track costs and use AI responsibly.

Future Outlook

Legal AI is beginning to look less like traditional software and more like electricity or cloud computing: you pay for what you use rather than simply for access.

If that trend continues, law firms may stop thinking about buying software licences altogether. Instead, they will be paying for computing power that helps lawyers complete legal work. That would change how firms budget for technology, price their services and decide where AI delivers the greatest value.