TL;DR
The Law Society, which represents solicitors in England and Wales, published a foresight study on Monday 5 October arguing that AI agents, which act on a user’s behalf rather than just generate text, could become the “default doorway” through which people meet the justice system. It wants accountability, transparency, “meaningful human oversight” and a public debate settled before such systems spread.
What the report predicts
The study looks a decade ahead. Its authors spoke to around 30 experts across law, technology, academia, the judiciary and regulation, and looked at examples from the UK, the US, China and Estonia.
The central forecast is a reversal of roles. Today, judges and lawyers decide with machines helping; tomorrow, machines may act while humans supervise. Agents could spot a legal problem before the person affected notices it, draft the claim, negotiate and chase the institutions involved. That raises a hard question about blame when an outcome is the product of several people and several automated systems acting in sequence.
The report also suggests the real gatekeepers may end up inside commercial platforms rather than courts or firms. If so, it warns, access to justice “could become shaped by private incentives, business models and invisible design choices”.
Not everything is up for automation. Deciding contested facts, judging credibility and exercising discretion are described as “constitutive of justice itself”, so removing the human there would damage legitimacy even if accuracy held. One of the few places the report sees a clean match with what courts need is routine, high-volume work such as sorting cases, booking hearings, processing documents and carrying out orders after judgment.
The pressure behind it
The Law Society frames this against a stretched system, with some court hearings already booked for 2030. It also notes that people are leaning on AI tools to find their way through proceedings, and that wrong or misleading answers from those tools are generating extra work for the courts. None of this is live yet: according to Legal Futures, the report accepts that no agentic system is working inside the justice system today, only that the groundwork for one is being laid.
Vice-president Brett Dixon was blunt: “Legal judgement cannot be delegated.” He added that AI “is not a replacement for much-needed investment in the court estate and additional court staff”.
Looking forward
The report’s sharpest line may be its warning that accountability “may arrive after harm rather than before it”. In our view, that is the practical message for UK law firms: anyone building or buying agent tools for client work should decide now who signs off each automated step, because the report expects regulation to trail the technology.