Bloomberg's story on how top law firms are using AI to transform billing, hiring and legal work reveals something most UK firms have not yet faced: the billable hour is no longer the default economic model. When Harvey, Luminance and other legal AI tools begin to complete research, due diligence and contract review in hours rather than days, the traditional six-minute increment collapses. This matters intensely to mid-market UK law firms because you sit between the magic circle (which can afford to absorb AI investment losses) and the high-street practices (which cannot yet afford premium AI at all). The story exposes a real tension: your biggest clients are now expecting AI-accelerated delivery. Your profit model assumes traditional timekeeping. One of these has to give.

What Bloomberg is documenting is the second wave of legal AI adoption. The first wave—document automation, contract assembly, basic legal research—was absorbed into existing workflows without fundamentally changing how firms charge. This second wave is different because it directly improves the commodity part of legal work: the grinding, billable research and review that junior lawyers and paralegals traditionally did. When a firm can run a 10,000-document disclosure review through Sift or Luminance in days instead of weeks, the economics of that work shift permanently. Clients will not pay traditional rates for accelerated output. Firms that depend on volume billing of junior-level work face margin compression, hiring pressure and (paradoxically) the need to invest heavily in senior expertise to stay differentiated. This is reshaping hiring, pricing and organisational structure simultaneously. It is not a technology problem; it is a business model problem.

Here is where many firms are going wrong: they are treating AI as a productivity multiplier within the existing billable hour framework. Harvey and Copilot are positioned as fee-earner tools that make lawyers faster. But speed within a per-hour model creates a dilemma—faster delivery cuts billable hours. The firms succeeding at this transition are not using AI to accelerate billable work; they are using it to change what they charge for. They are moving toward value-based pricing, fixed fees and outcomes-based models. That requires two things: transparency on how AI improves cost and risk, and robust governance to prove to the SRA and your clients that the work quality has not suffered. Trovix Audit exists precisely because AI governance is where most firms stumble. You cannot shift to fixed-fee delivery using unauditable black-box tools—clients will not accept it and regulators should not either. The FCA's Consumer Duty PS22/9 and the SRA Code require you to show that cost savings benefit clients, not just your firm. If you cannot demonstrate that, you are naked to competition and complaint.

What to do now: mid-market firms need to make three decisions this year. First, audit which parts of your workflow are genuinely being disrupted by AI (hint: it is not the specialist advice, it is the commodity search and assembly). Second, decide whether you will chase volume-based efficiency or shift toward outcomes-based pricing for those commoditised tasks. This is not a technical decision—it is a commercial one, and it requires partnership conversations with your biggest clients. Third, implement AI governance and cost transparency tooling before you scale AI deployment. You cannot move to fixed fees on unauditable systems. Trovix Reach can help you communicate what AI does (and does not do) to clients. But the decision to change your model must come first, and the AI tools must follow it, not the other way around.

Source: Bloomberg News