AI law firm Garfield AI has won what is believed to be the first court case prepared almost entirely by artificial intelligence, securing a £7,000 judgment at Wandsworth County Court on May 14. Human lawyer Dominic Li argued the case, but the pretrial work-document drafting, witness statements, and counterclaim responses-was all handled by the firm's AI software. The outcome signals a growing push to use AI to slash legal fees and make enforcement of smaller claims financially viable.
The pretrial AI edge
Freelance HR consultant Tamires Camal Taquidir paid roughly £400 to send a legal letter through Garfield AI after a client failed to pay a £7,000 debt. The AI then took over the pretrial litigation, generating the court bundle, preparing four witness statements, and disputing a counterclaim the defendant raised. "When the counterclaim was brought, it was intended to intimidate me, but I knew I had accessible, cost-effective, and competent support. I'm delighted by the result," Taquidir told The Guardian.
Li, who presented the case in court, said the AI-prepared materials were "clearly and efficiently" assembled. Yet he stressed that human advocacy remains essential. "The advocacy at trial remained essential and a fundamentally human exercise," he told the same outlet. For paralegals, draft witness statements and document bundling are central to trial prep-tasks an AI Learning Path for Paralegals can help professionals approach with generative tools while maintaining oversight.
Where AI still stumbles
Large language models are prone to hallucinations-inventing facts and citations that look plausible but don't exist. In September 2025, a California attorney was fined $10,000 for citing non-existent court cases generated by AI, as Reuters reported. The following April, prosecutors in Nevada County acknowledged errors in four criminal cases after using AI to draft and research legal briefs.
District Attorney Jesse Wilson wrote in a brief to the California 3rd District Court of Appeal: "The office acknowledges that it was not fully prepared for the emerging risks presented by generative AI, including the prevalence of its use, the extent to which such tools may affect the accuracy of legal work product, and the difficulty of detecting deceptively plausible fabrications without careful scrutiny."
Despite these risks, California's largest courts are testing an AI tool to assist with criminal case decisions, including appeals involving racial bias. Legal experts believe the technology could ease backlogs; opponents warn it will dehumanize justice and amplify hallucinated citations. Law firms adopting these tools need rigorous verification processes-a skills gap that AI for Legal Professionals Courses aim to close by teaching structured output checking and ethical safeguards.
Why this matters for legal professionals
The Garfield AI win proves that routine pretrial tasks-document drafting, bundling, and simple motions-can be offloaded to AI, freeing lawyers for courtroom advocacy and strategy. But the string of hallucination-related sanctions makes one thing clear: no firm can afford to treat AI output as final. Every generated document must be verified, and practitioners need to understand the technology's failure modes. Professionals who can couple AI productivity with rigorous human judgment will be best positioned as courts experiment with these tools. For those in litigation support roles, learning to direct and audit legal AI systems is fast becoming a core competency rather than an optional skill.
Your membership also unlocks:
