AI agents could one day guide people through legal problems, draft claims and negotiate with institutions, the Law Society warns. Its foresight study argues that accountability and meaningful human oversight need settling before such systems become a doorway into justice.
Watch Desk analysis
What happened
The study looks a decade ahead and draws on around 30 experts across law, technology, academia, the judiciary and regulation. As described by Resultsense, it imagines a shift from lawyers and judges using machines as tools to agents taking action while people supervise.
The report says no agentic system is currently working inside the justice system. It identifies routine, high-volume work, such as sorting cases, booking hearings and processing documents, as a better fit for automation than deciding contested facts, judging credibility or exercising discretion. The Law Society’s foresight study also raises the prospect that commercial platforms, rather than courts or law firms, could become gatekeepers to legal help.
Why it matters
The stakes are practical as well as philosophical. If an agent takes several steps on someone’s behalf, responsibility may be hard to pin down when something goes wrong. The report warns that accountability could arrive after harm, while people already using AI tools to navigate proceedings can bring misleading answers into an overburdened court system.
The Law Society sets this against long waits, with some hearings already booked for 2030. Automation could help with administration, but it cannot substitute for investment in court staff and buildings, vice-president Brett Dixon says.
Our read
This is a warning about the rules being built around a real prospect, not evidence that AI agents have already entered courtrooms. The distinction matters. Automating a hearing slot is not the same as automating a judgement, and the report is right to draw a firm line between the two.
For law firms considering agents in client work, decide who approves each automated step before the technology makes that an awkward question to answer. A human in the loop is useful; a human who cannot explain what the loop did is less reassuring.
What to watch
- Whether the Law Society turns its principles into specific proposals for agent accountability and oversight.
- Which routine court processes are considered for automation, and what safeguards accompany them.
- Whether future systems disclose when an agent has acted and who is responsible for its decisions.
Discussion spark: Should AI agents ever be allowed to negotiate or file claims on someone’s behalf, or should their role in justice be limited to administrative tasks?
Sources and evidence
- Law Society warns the justice system is not ready for AI agents – Resultsense (6 October 2026, 07:25 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.