AI is already changing how people prepare legal claims and how judges handle information. Geoffrey Vos, who is about to retire as Master of the Rolls, believes it could eventually help decide some simpler civil disputes. The idea raises a central question: how can courts gain speed and access without undermining confidence in justice?
From paperwork support to decisions
In England, current guidance advises judges to limit AI use to administrative work. Vos says he uses Microsoft Copilot to locate older documents, find information, navigate large collections of material, and check details. He says he does not use it to write judgments.
That distinction matters. Finding a relevant document can save time, while deciding a case requires weighing evidence and reaching a result that affects the people involved. Vos says he is more cautious about using AI to read papers in cases because he handles some of the system’s most important matters.
His view was shaped in part by a visit to a court in Shenzhen, China, where civil disputes are decided with heavy assistance from AI. He was impressed by what he saw and thinks courts elsewhere should consider whether similar approaches could work.
More people are bringing claims
Vos says county court cases have risen by 10 percent each year, and he thinks that increase may continue. People with legal problems are turning to Gemini, ChatGPT, or Claude to draft claim documents and forms.
The quality of those claims is not clear. Vos says there is no data showing how many are based on sound legal principles. He cautions against assuming that a claim prepared with AI is frivolous: a person’s initial description may not capture the legal issue precisely, but the underlying problem may still deserve attention.
For Vos, this rise can also reflect wider access to justice. Legal help has historically been more available to people with money. AI tools may make it easier for others to bring a concern forward, even if courts then need to examine what the claim actually involves.
Trust depends on how courts use AI
AI use in legal work has also created clear risks. Lawyers have been caught citing case law invented by AI, and one person in the UK recently lost their license to practice in part because they misused the technology. Chatbots have also helped generate claims, including some that are bogus.
Vos argues that courts should explain how judges adopt new technology and build broad agreement that its use is appropriate. Confidence in the justice system, he says, rests on people believing judges act with integrity. If a process is opaque, the public may question whether a decision was reached fairly.
There is also the question of fabricated evidence. AI can produce false images, video, or voice recordings. Vos sees these as an extension of a longstanding judicial task: deciding whether evidence is genuine and whether a person is telling the truth. He points to forged wills as an older example of false material courts have had to assess.
He is optimistic, but says people who submit fake images to court can face the consequences of perjury. That risk does not remove the challenge of assessing evidence; it underscores why courts must keep examining what is presented to them.
Where machine decisions might fit
Vos sees a possible role for AI in relatively straightforward civil matters, such as personal injury claims and low-value professional negligence cases. He says complicated commercial disputes and sensitive family cases may still call for a human judge.
Any move toward algorithmic decisions would require public acceptance. Vos says people would need to consent to the process in a particular case, or society would need to decide through legislation that it is acceptable. Even when people dislike a human judge’s outcome, they may come to accept it over time; a decision attributed to a machine could be harder to accept.
The prospect, then, is not simply about whether AI can produce an answer. Courts would need to decide which cases are suitable, how to explain the process, and how to maintain trust when someone receives an unfavorable result. Vos’s argument is that justice must keep pace with how people use technology, while approaching the change cautiously and openly.