The hype is getting old, or at least is becoming exhausting. First, widespread experimentation. Then the crisis. Then the cautious rebuilding. Legal AI has hit the point where results matter more than promises, writes Ryan Zahrai.
You could see it coming. By early 2025, global firms were reporting system failures, erratic outputs, and answers that simply didn’t hold up. The legal sector, which had spent months posting about “AI assistants” and “AI copilots”, suddenly realised those tools needed guardrails. By Q3, things stabilised. Some lawyers were using AI daily and loving it. Others shut it down entirely after a few bad experiences. Security concerns, hallucinations with case law citations, chastisement from courts, accuracy issues, and a lack of trust kept a large portion of the market on the sidelines. The divide was real.
Now, heading into 2026, the mood is more grounded. The experiments are behind us. For the more tech-enabled among us, the novelty has worn off. Firms want proof that a product works, is accurate, materially drives efficiency, thereby increasing margins and operational speed to delivery, and a reason to trust the tech. They want tools that earn their place in a workflow. Everything else is noise.
AI has hit the legal sector hard, but at least from a cost perspective, clients have not yet realised any material benefit. The data suggests that firms now use AI in one way or another. We can only assume that efficiency is up and internal cost is down, yet pricing looks the same. That won’t last.
The anchor is time-based billing. Even with AI cutting hours off research, drafting, and review, many firms still charge as though nothing changed. It creates a short-term margin lift but ignores what is coming next.
In-house teams already expect AI-driven pricing shifts, and the surveys confirm it. When the demand side gets informed, billing models change. Other industries have started to make that shift. While ours may be slower to adopt, and traditionalists will do what they do best (i.e. resist change), the legal profession won’t be the exception.
That tension is exactly why we shifted our own model. We’re a micro team at my firm, but because we built an AI-first, human-verified workflow from day one, we operate at three times our size. Routine work moves through our internal and client-facing systems so our lawyers can focus on judgement, commercial steering, negotiation, and risk assessment. Those tools aren’t proprietary to us – just the flow and way we’ve integrated them is unique.
That gave us room to price the work on value rather than hours. Now, where we can, we don’t charge for the time AI saves us. We charge for the outcome. Clients pay for what matters, not the inefficiency the system used to hide. They get speed without footing the bill for routine tasks that no longer take hours. That means that, on many occasions, clients pay us less than what they’d pay another firm, but because we’re not billing for time, we’re making margins on a deliverable that we wouldn’t be able to make if we had to charge on a time basis. That’s a win-win.
What we expect in 2026
Based on profession data and our own experience, the next year will be shaped by four clear trends.
The shift in business models
There are three forces that will reshape legal services in 2026.
The first is margin expansion. Firms that deploy AI properly will reduce internal costs and increase delivery speed. That creates margin, but only while the pricing model remains tied to time.
The second is pressure from clients. As more industries adjust pricing to reflect AI efficiency, law firms will be expected to do the same. Clients will shop for firms that can deliver routine work faster and cheaper, but at the same time without compromising on quality and expertise.
The third is the move towards recurring revenue. AI makes advisory subscriptions, fixed retainers, and embedded legal support commercially viable. These models align incentives. Clients get predictability and flat legal opex P&L lines in their P&Ls. Firms get compounding margin, and the competitive advantage of being AI-ready.
The winner’s circle
The firms that win the next decade will not be the ones with the most AI features. They will be the ones that combine AI-first delivery for routine work, strong human capability for complexity, predictable value-based pricing, and a recurring revenue model.
Clients will not pay premium rates for machine-assisted work. They will pay premium rates for human-in-the-loop judgement and the ability to manage complexity.
The opportunity for the broader profession is obvious. Use AI to increase margin. Use modern pricing to pass some of that benefit to clients. Do both and you end up with a stronger, more defensible, future-proofed business.
Ryan Zahrai is the founder and principal of Zed Law.
