Boonchai Wedmakawand | Moment | Getty Images

Artificial intelligence is now used by almost 90% of legal professionals in the U.K. and Ireland, and it's putting one of the profession's oldest conventions — the billable hour — under the microscope. That's according to legal software company Clio's U.K. & Ireland Legal Insights Report 2026. It found that among firms using AI, almost 80% said they can handle more work without increasing resources, while over 70% said it cut costs by absorbing administrative work once done by support staff. As a result, AI is challenging some of the assumptions on which the legal profession was built, forcing firms to reevaluate how their lawyers spend their time, how they charge for it and how new lawyers learn the ropes.

You can't charge 16 hours for something that takes 16 seconds Nick Rowles-Davies Lexolent

Some of the U.K.'s biggest firms are already putting this into practice. A&O Shearman has worked with legal AI company Harvey to develop artificial intelligence agents for tasks, including reviewing loan agreements and analyzing regulatory filings, which it says can complete in minutes work that previously took several hours. Slaughter and May, meanwhile, has rolled out Harvey across all practice areas this year, including for regulatory research and document analysis.

Billable hour pressure

The billable hour is central to the business model of many law firms, but when AI significantly reduces the time lawyers spend sifting through and drafting documents, the economics are no longer so straightforward. "You can't charge 16 hours for something that takes 16 seconds," Nick Rowles-Davies, founder and CEO of legal finance fund Lexolent, based in London and Dubai, told CNBC.

About one in five firms that have widely adopted AI report difficulty meeting billable-hour targets, according to Clio's report. Globally, senior legal leaders expect the share of work charged by the hour to fall from 72% to 44% over the next two to three years, according to a Deloitte survey. Routine work is the most exposed, Rowles-Davies said. "If you've got standard documents and you're just putting in detail, then clearly that's an automatic process." But complex legal work still requires human judgment, he added, particularly when interpreting AI output and determining the right strategy for a client.

Lawyers [are] telling us that their day is getting better Joshua Lenon Clio

Lawyers also need to be able to spot AI hallucinations. "There have been a lot of cases where lawyers have been sanctioned for producing briefs before a court with cases that don't exist," Rowles-Davies said. Mansoor Soomro, who leads the Future of Work Research Unit at Teesside University in Middlesbrough, England, told CNBC the billable hour is certainly not disappearing overnight, but it is getting harder to justify.

Junior lawyer training

AI is also changing how lawyers learn their trade, Soomro added, as repetitive tasks traditionally helped junior lawyers learn the building blocks of law. Reviewing hundreds of documents taught them what matters, writing up contracts honed their drafting skills and researching dozens of cases helped them develop legal reasoning, he added.

Wera Rodsawang | Moment | Getty Images

As AI takes on more of that work, firms will need to be deliberate about replacing those learning opportunities, according to London-based Jorge Bestard, head of EMEA at Harvey. "A faster route to a draft does not automatically mean a faster route to competence," he told CNBC. A&O Shearman is placing more value on "curiosity and entrepreneurship," said Francesca Bennetts, the firm's London-based head of legal talent. It is also building career paths in legal engineering, product and AI delivery that "barely existed a few years ago," she told CNBC. "The need to develop exceptional lawyers doesn't go away," Bennetts said. "They are the partners and experts of the future." Soomro expects judgment to become "disproportionately valuable" in the sector, with lawyers spending more time checking and challenging AI's output and interpreting what it means for a client. "The future may therefore not belong to the lawyer who knows the most about one narrow area, but to the lawyer who knows how to connect multiple areas to solve a complex client problem," he said.

AI and workloads