Autonomous AI 'agents' could tackle many of the justice system’s existing problems - but the risks posed by their development and use require tougher regulation than ordinary generative systems. That is the main thrust of the Law Society's latest 'foresight' report, which covers the rapidly developing world of agentic AI.
Agents that can take action on behalf of users could help reduce court delays and improve access to justice, the report states. But without robust safeguards, errors can cascade quickly, accountability can be fragmented and unclear or unequal access to advanced AI tools could widen inequalities.
The Future of agentic AI in the justice system draws on interviews with practitioners across the legal profession, technology, the judiciary, academia and policy, alongside member discussions and desk research.
Interviewees generally agreed that the administrative side of the justice system presents clear opportunities. Agentic AI could transform back office machinery into an 'active intelligence' resource, capable of predicting demand, prioritising cases, allocating resources and determining pathways through the process.
The use of agents for judicial decision-making is a more divisive topic. However the report notes examples of how functions once considered judicial are already being reclassified as administrative, included uncontested divorce proceedings, court listing, immigration and asylum chamber decisions.
In the future, the report states: 'The most influential decisions about how justice operates may no longer happen in courtrooms, but inside automated systems that few people see, understand, or can challenge.'
However a 'consistent theme' among interviewees was the gap between the pace of technological development and the development of regulatory frameworks. 'Interviewees were clear that agentic AI requires a more demanding regulatory response than generative AI. Because agentic AI systems act autonomously across multiple steps rather than responding to individual prompts, an error at any point can feed into the next step before anyone notices.'
Some interviewees argued that agentic AI needs its own regulatory tier rather than being treated as a more powerful version of existing AI tools.
Society vice president Brett Dixon said: AI is not a silver bullet to improve the justice system. It could help to ease some administrative pressures, but it is not a replacement for much-needed investment in the court estate and additional court staff. Legal judgement cannot be delegated. Not in courts and not in the legal sector. Professional expertise that shapes, checks, questions, improves and approves AI outputs is needed now more than ever.
'The most important question becomes not what AI can do, but what should remain human. There is still time to shape the boundaries of AI before the market does it for us.'
No comments yet