Excitement and anxiety about how artificial intelligence will change law took centre stage last week at one of the world’s largest legal technology conferences.
More than 5,000 lawyers, technology suppliers and consultants gathered at the annual LegalTechTalk, in London’s Greenwich peninsula.
The atmosphere was upbeat, scattershot and slightly frazzled, as attendees reflected on the four-year, AI-powered boom in legaltech and what it could mean for their jobs and clients.
Law firms and company in-house legal teams are increasingly using AI to save time and money by automating tasks such as research, drafting and contract review, while also making AI integral to their strategies.
Companies will double their legal technology budgets by 2028 as they attempt to build on early gains in productivity and efficiency from AI-powered legal software, Gartner has forecast.
Whether that will be money well spent is currently being debated.
AI optimists, predict that legal AI software will automate boring but essential legal tasks, such as legal research and comparing lengthy contracts. This, they argue, will allow lawyers of all ranks to spend more time on more interesting and higher-margin work, such as strategic advice and winning new business, and tasks that require more judgement and experience.
AI sceptics argue that the boom in legal AI is a bubble with too many suppliers in the market to be sustainable. They also note the risks of legal AI making stuff up or “hallucinating” − adding that many law firms and company legal departments are still in the experimental phase of AI and, so far, are only able to show modest payback from the technology.
Max Junestrand, co-founder and CEO of Legora, a Swedish legal AI start-up valued recently at $5.6bn, is, unsurprisingly, among the optimists.
In a conference panel discussion, he reflected on the rapid advances in AI functionality in the past nine months. “If you walk the hall of Legora [there] are engineers literally whispering into microphones telling agents how to code,” he said. “[The] same things are happening in law.”
Legora is including agentic AI – software capable of performing tasks without human intervention − into its products.
“At Legora, we like to talk about the agentic hamburger – you need different layers to complete a gourmet meal,” Junestrand said.
Using advanced AI, is rarely a simple matter of plug-and-play, however, as other tech suppliers and law firms speaking at the conference acknowledged.
The value an organisation gets from AI will depend on the quality of information fed to the AI software and whether its lawyers have the skills and inclination to use the technology, conference speakers said.
“[The] most common error I see people make… [is] still operating the same processes and they try to add an AI layer on top of it [and] use this to reinvent stuff,” said Sarvarth Misra, CEO of Leah, which makes agentic AI software for industries including legal, financial services and retail. “If you take a crap process and put AI on top of it, it will still be crap.”
Among law firms at the conference, there was enthusiasm for AI, but also signs of fatigue at the sheer number of legaltech suppliers to pick from and new products to pilot.
Nicola Shaver, CEO of Legaltech Hub, a data provider and adviser on legaltech, estimates that there are more than 1,000 generative AI products in the legal sector.
“Suppliers should be honest about their roadmaps,” said Hélder Santos, global head of legal tech and innovation at international law firm Bird & Bird, in a conference panel discussion. “Forget the [demonstration] theatre – everything works fine in the demo.”
At Bird & Bird, lawyers are not given targets for the number of times they should use AI, Santos said. “I don’t care if the best lawyer uses only three or four prompts if one of them brings the firm to the next level.”
And although the firm’s lawyers are advised which AI tools to use at work, they are also allowed to experiment with non-authorised AI tools in digital “sandboxes” − an isolated digital environment where new technology can be tested safely − he said.
In a panel discussion, former GE Vernova general counsel Rachel Gonzalez said the human factor in AI is probably the biggest challenge for legal teams.
The “velocity” of change in AI means “getting everyone on the same page” is challenging, she said.
Although Gonzalez was generally positive about AI’s likely impact on the legal sector, one potential concern is an increased “cognitive load” on lawyers if AI automates routine legal work, leaving lawyers spending all their time on more complex, high-value and high-pressure legal work, she said.
AI’s impact on legal jobs was another talking point. There have been predictions that AI will automate away swathes of white-collar jobs, including in law, and make it hard for entry-level lawyers to learn legal basics, such as drafting and reviewing a contract.
At the legaltech conference, however, there was no talk, in public at least, about AI replacing human lawyers and pulling up career ladders.
Alison Malin Zoellner, group GC at advertising company Dentsu Group, said her junior legal team’s aptitude for AI was valued at the company, including in advising more senior staff on how to use AI. “A lot of reverse mentoring is going on right now,” she said.
Elsewhere in the conference, there was discussion about how traditional law firms respond to challenges posed by AI. Several speakers mentioned “the innovator’s dilemma” concept, coined by management expert and author Clayton Christensen, about how incumbent companies can miss out on new and innovative technologies and be usurped by smaller companies.
One such challenge is the loss of lawyers to so-called ‘native AI’ law firms, such as Norm Law LLP, that are built from the ground up using AI.
Shaver advised law firms to develop their own AI, work out what their “superpower” is, who are their most important clients and then embed AI in those areas. “Adoption of third-party AI products is not enough… that’s table stakes… a new shape of market is coming to eat them.”
Email your news and story ideas to: [email protected]
