The solicitors' watchdog has recently cleared an artificial intelligence law firm to offer legal services through AI rather than through lawyers. (iStock.com/Tero Vesalainen)
Garfield's co-founder, Philip Young, told Law360 that his firm aims to help businesses and tradespeople collect small debts through the county courts. But he added that he wants to offer more.
"We'll definitely build and deliver more products because I really want this to help people," the former City commercial litigator said. "It will make it easier and cheaper to access the justice system in all kinds of different cases."
Small businesses fail to recover billions of pounds in unpaid debts every year, often because it is too costly or time-consuming to pursue claims under £10,000 ($13,500).
"The problem we are trying to solve is enormous," Young said. "A lot of people who are unrepresented at the moment are turning to consumer-grade AI like ChatGPT and trying to use those to help them. And that's unsatisfactory and dangerous."
Garfield.Law, which is powered entirely by an artificial intelligence large language model, and the other AI-driven law firms that the SRA expects will emerge have considerable promise. Justice Colin Birss described Garfield in November as "absolutely at the core of what we can do for access to justice."
A recent survey commissioned by a legal technology startup found that only 10% of people fully trust law firms but that almost one in three would be willing to allow a "robot" lawyer to represent them in court. But almost all would only do so with a human lawyer overseeing the process.
As the firm's designated compliance officer, Young oversees all aspects of its operations, including the AI-generated letters that it uses to pursue debtors or initiate legal action through the small claims court process. "In the early days, I'm checking everything, all the documents that are going out," he said.
Ultimately, however, he expects to shift to a sampling process concentrated on higher-risk work. Young said that the technology should be able to reliably assess its own accuracy once it has sufficient data, reducing the need for routine human oversight.
"You can use technology as a first sift to ensure quality, and then bring in humans only when there's a problem or a question mark that needs to be analyzed," he said.
Julie Bishop, director of the Law Centres Network, a representative body, compared the AI-powered software platform used by Garfield to supermarket self-checkouts — useful, but only when support is readily available.
"What made it more popular and widely used is that there's always someone nearby to help when it doesn't work properly," Bishop said.
But she warned that Garfield's current services would not suit the law centers that support individuals facing problems such as eviction because they haven't been able to pay their rent. For such organizations, any new tools must include safeguards and be "cost-effective and reliable," allowing them to meet the needs of vulnerable clients within tight budget constraints.
But, Bishop added, "If these tools can help people who are able to help themselves, for us at law centers that's good news because it frees us up to help those who really need to be sitting with someone by their side."
Law centers prioritize efficiency. They use tools such as Microsoft 365 Copilot, Wyser and Contend Legal to save lawyers time and offer better support to clients in face-to-face settings.
Citizens Advice, a network of local independent charities that provides free legal advice in areas such as managing debt, consumer rights to housing, takes the same approach.
Stuart Pearson, head of innovation at several local Citizens Advice centers, said that although technology such as that used by Garfield.Law might not directly apply to a charity model it reflects the growing potential of AI to support legal and advice services.
Its core features — advanced information retrieval, analysis and support for advisers — could be "incredibly valuable," he said. It has the potential to streamline processes, manage information more effectively and free up human advisers to focus on their clients' often-complex needs.
This kind of innovation is on Citizens Advice's radar, Pearson said. His team has successfully trialed a Gen-AI tool called Caddy, which has halved average phone call times from around eight minutes to four.
The AI assistant, which is set to be rolled out nationally this year, helps call handlers to quickly find answers using approved internal sources. Caddy is replacing slower methods including manual searches or checks with colleagues, which often require handlers to put callers on hold.
The new tool "means clients get the information they need faster, and our advisers can help more people," Pearson said. "Our focus with tools like Caddy is firmly on using AI to amplify the skills of our human advisers, not replace them."
Young of Garfield.Law does not want to see his technology replace lawyers. "I've got a vested interest in that because I am a lawyer and I don't want to replace myself," he said.
He said that was why Garfield.Law was built for law firms too. "What we offer is basically something that will enable them to benefit from the ability to do more claims without having to take on significantly more staff," he said.
He argued that this also reduces the need for additional office space and overheads, allowing firms to operate more efficiently and improve profit margins.
Law firm consultant Viv Williams said that some cases such as complex probate matters or conveyancing involving large properties or questions about brand registry will still require human oversight. But he said that almost all of this type of legal work will eventually be handled by AI.
This transformation "will have a dramatic effect on the way legal services are delivered in the next few years." Williams continued. "It's only a matter of time before the profession accepts it and moves on."
Williams warned that firms such as Garfield.Law could put pressure on small practices that struggle to compete on cost. And sole practitioners are hesitant to adopt new technology, he added.
"I'm dealing with several law firms at the moment where the owners are in their 60s and 70s, and you think, 'Well, are you really going to adapt to AI?'" he said.
But it will also affect entire teams within traditional outfits. "If you have a law firm with a department that specializes in debt recovery, they're competing now against a computer, effectively," Williams added.
--Editing by Ed Harris.
For a reprint of this article, please contact reprints@law360.com.
