A doctoral student at the University of Oxford successfully defeated a British energy company's three-year debt collection effort for a non-existent address by using AI tools to build his legal case, according to a report published by The Register on August 31. Lyle Hopkins, a software engineer pursuing his PhD at Oxford Interdisciplinary Bioscience DTP, turned to GPT-5.5 and Claude Fable for legal guidance after SSE Energy Supply billed him £1,091.01 over more than 20 months for an unused electricity meter tied to a previous owner's failed flat conversion. Representing himself in court, he won £1,087.88 in compensation, expenses, and interest on July 17 when a judge ruled in his favor at the County Court at Oxford.
SSE Energy Supply pursued Hopkins for business-rate charges connected to a disconnected meter at unit 8b, an address that never existed at his property, the report details. A company representative sent an email on June 15, 2024, confirming Hopkins bore no liability, yet SSE subsequently dispatched debt collectors to recover the amount. Hopkins never paid the bill but wanted to stop the harassment. He spent approximately £175 on three active AI subscriptions and additional API fees, describing the investment as "painful" because of tight deadlines that forced him to pay API rates. Before the hearing, SSE offered Hopkins more than the court ultimately awarded, conditional on confidentiality and non-disparagement clauses, which he refused.
"I used AI to test an energy company's court claims against its own records, then ran the hearing myself and won," Hopkins said. According to Hopkins, he wouldn't have been able to cite the case law without AI assistance, though he had to review everything and ensure the case law actually existed. The judge hearing the case said Hopkins, who sought help from an Energy Ombudsman after repeated communication with SSE, "could not have done more to make clear that he was not liable for claimed debt." The judge added that "the defendant has subjected Mr Hopkins to a rollercoaster ride and not a good one," warning that continuing to pursue Hopkins for the liability would constitute harassment.
Hopkins didn't consider hiring a solicitor because the cost would have been prohibitive, while his AI spending remained relatively modest despite burning through substantial tokens across his Github Copilot, OpenAI, and Anthropic subscriptions. The report notes that when Hopkins first submitted his claim on October 7, 2025, GPT models hallucinated more frequently and weren't as effective at checking case law, and if Claude Fable had been available then, he probably would have claimed around £5,000 and included harassment charges based on the case law it later found. Hopkins told The Register the case dragged on for nearly three years and cost hundreds of hours, lost sleep, and time that should have gone into his medical-research doctorate at Oxford, with his exhibits alone exceeding 100 pages. He attributes a delay in completing his doctoral degree to time spent on the case, noting his funding ended before he could finish, and has launched a funding campaign to support completion of his doctorate.
Even after Hopkins won his case, SSE sent another bill dated July 23 demanding payment, six days after the July 17 judgment that ruled further collection attempts would constitute harassment. Hopkins remains uncertain whether he'll pursue a harassment claim, expressing concern that people less academically equipped than himself would end up with county court judgments and ruined credit ratings over debts they never owed. An SSE spokesperson told The Register after the story was filed, "We apologise for any inconvenience caused to Mr Hopkins during this process. Following the Court's judgment, we have paid compensation to the claimant." The case demonstrates how accessible AI legal tools could reshape power dynamics in small claims disputes, though it also highlights the persistent need for human oversight to verify AI-generated legal arguments against actual case law. For organizations facing similar consumer challenges, the incident serves as a reminder that automated billing systems require rigorous human review before escalating disputes to collection or legal action.