British AI Law Firm Wins Its First Court Case

Freelancer Tamires Camal Taquidir recovered £7,000 in unpaid wages from a catering company with the help of Garfield AI, a British law firm built on artificial intelligence. The case, decided in May 2026 before Wandsworth County Court in London, is being described as the first instance in which a court case was won by a firm built around an AI model rather than a traditional law firm.

Garfield AI was founded by Philip Young, a former Baker McKenzie lawyer, and Daniel Long, a quantum physicist. The company describes itself as the world's first law firm operating entirely on artificial intelligence. Its business model focuses on small claims and debt recovery within the UK's small claims court system, where the value of a dispute is usually too low to justify hiring a traditional lawyer.

How the AI ran the case

The Garfield platform guided the client through the entire pre-trial process, from the letter before claim through drafting the lawsuit to compiling the evidence and witness statements needed for the hearing. The AI system did not, however, replace a lawyer in the courtroom. To represent the client directly before the court, the firm brought in a professional barrister, Dominic Li of One Essex Court.

The defendant filed a counterclaim, which the court dismissed, ruling entirely in the claimant's favor. Barrister Li judged the AI-prepared documentation fully sufficient to run the case. Notably, the opposing side had both a solicitor and a barrister, yet still lost to a client represented by the cheaper AI-based model.

This is the first case an AI lawyer has ever won against a human opposing party - Philip Young, co-founder of Garfield AI

Regulated AI-supported legal services can help ordinary people recover real money in court - Daniel Long, co-founder of Garfield AI

Regulation as the key to trust

A key element of this story is Garfield AI's regulatory status. In 2025, the UK's Solicitors Regulation Authority (SRA), the body overseeing the legal profession in England and Wales, authorized Garfield.Law as the first firm to provide regulated legal services built on artificial intelligence. That means the company is held to the same oversight standards as traditional law firms, setting it apart from the many AI tools that offer only generic legal advice with no professional accountability.

This status has let Garfield AI build a genuine track record of cases and outcomes it can point to as evidence the model works. More than 600 claims handled and roughly half a million pounds recovered for clients are numbers hard to ignore in the legal industry, especially since the cost of representing the client in this case amounted to just under six percent of the sum she ultimately recovered.

What it means for legal services

Small claims have long been an area where traditional law firms are unwilling or unable to operate profitably, since legal costs outweigh the value of the dispute. Garfield AI's model, pairing automated documentation with a single barrister engagement only at the hearing stage, drastically cuts costs and opens access to justice for people who previously couldn't afford it.

Industry observers note that if a similar model proves repeatable in other cases, it could put pressure on traditional firms handling low-value disputes and prompt other jurisdictions to develop similar regulatory frameworks. In the UK, Garfield AI's success could also speed up the SRA's decisions on approving further AI-based law firms.

For the Polish legal market, which still has no equivalent of a UK-style small claims court run this way, the Garfield AI case points to the direction AI-supported legal services regulation could take. Polish law firms and legal startups are increasingly testing AI tools for drafting court filings, but none has yet obtained the status of a regulated legal services provider comparable to the British model.

The case unfolded amid growing criticism of uncritical AI use in the judiciary, including warnings from judges and regulators about representatives using chatbots without verifying their output. Garfield AI differs from those cases in that it operates as a supervised, regulated entity with a human lawyer responsible for the final stage of the process, rather than as a tool replacing professional legal judgment.