UK lawyers are poised to unlock £2.4 billion in productivity savings through artificial intelligence adoption, according to new research from Thomson Reuters. The study reveals that legal professionals expect AI tools to free up approximately 140 person-hours over the next year, translating to roughly £12,000 in productivity savings per lawyer.
The potential for AI-driven efficiency gains in the legal sector appears set to accelerate significantly. Within three years, lawyers anticipate AI will have saved them 240 working hours annually, equating to £4 billion in sector-wide savings. Looking ahead five years, the projection extends to 370 hours - equivalent to 15 working days - representing more than £6 billion in productivity gains.
Thomson Reuters points to its agentic AI solutions as exemplifying the next wave of sophisticated AI tools delivering substantial productivity improvements. These advanced systems can plan, reason, act, and react with the precision and auditability that legal workflows demand, without requiring continuous user prompting.
However, the research underscores a critical distinction between professional-grade and consumer AI tools. To maximise return on investment, Thomson Reuters emphasises that law firms need clear AI strategies built around expert-grounded, auditable, and workflow-integrated solutions designed specifically for high-stakes professional environments.
John Shatwell, head of Legal Professionals Europe at Thomson Reuters, warned against relying on public AI tools for professional legal work. 'Law firms that have a clear strategy centred around professional-grade AI will get results, while those that don't will waste millions on consumer orientated, public AI tools that simply can't handle a high-stakes professional environment,' he said.
The fundamental issue, according to Shatwell, lies in the training data and reliability standards. 'Public AI tools, which are trained on data scraped from the internet, break down under the trust, accuracy, and compliance standards demanded by legal professionals,' he explained.
To address these challenges, Thomson Reuters trains its AI tools on proprietary, trusted content sources and maintains a human-in-the-loop approach throughout the AI model lifecycle. The company employs over 1,500 trained lawyers to ensure the accuracy and currency of legal information underlying its AI solutions.
The research, based on data from 197,400 solicitors in the UK and average employment costs, suggests that while AI presents unprecedented opportunities for legal sector efficiency, success will largely depend on firms' strategic approach to AI adoption and their choice of professional-grade versus consumer tools.
