King's College London

The UK justice system faces mounting challenges, with significant delays in case processing causing extended waits for justice. One major reform introduced in 2015, the Single Justice Procedure, was designed to streamline legal proceedings by allowing a single magistrate to decide cases entirely through written evidence, bypassing traditional court appearances. While efficient in theory, this approach has raised concerns about fairness and accessibility.

For instance, a motorist accused of running a red light would receive a postal notification asking them to plead guilty or not guilty. Failure to respond, or choosing to plead guilty, results in sentencing by a single magistrate, without a physical court hearing. Data from HM Courts & Tribunals Service indicates that a high number of Single Justice Procedure cases lead to convictions in absentia (while not present). A report from Transform Justice highlights that vulnerable individuals are disproportionately affected, often failing to engage with the process for many reasons, including financial and social barriers. Recent cases have exposed inconsistencies and overlooked evidence, prompting a Parliamentary Justice Committee review, which flagged concerns about procedural fairness.

Advances in Artificial Intelligence hold promise for tackling inefficiencies and inconsistencies within justice systems, yet their role in legal decision-making remains largely untested and lacks public scrutiny. Our project seeks to bridge this gap, collaborating with justice reform advocates, including Transform Justice, to develop and rigorously assess AI-driven decision-making models. Through a structured research process which includes algorithm development, stakeholder workshops, and a cycle of testing, gathering feedback, making changes, and testing again to continuously improve the system, we will establish evidence-based frameworks that balance automation with ethical safeguards.

At the core of our approach is a "human-in-the-loop" methodology, ensuring that AI enhances judicial decision-making rather than replaces it. By integrating stakeholder perspectives, refining models through expert review, and prioritizing fairness, transparency, and accountability, we aim to harness AI for justice, not just efficiency.

“It means a lot to receive this award. There is a great deal of important work to be done around the question of how AI and algorithms impact justice. By working with an incredible team of partners from King's and beyond, we hope to provide insights and techniques that will be useful for the public, for policymakers, and the research community. ”

- Cari Hyde-Vaamonde, PhD Student, Lawyer and GTA in Law, and Dr Pierpaolo Vivo, Reader in Disordered Systems in the Department of Mathematics

Expected Impact

The project will:

• Develop a working prototype AI tool to assist magistrates with traffic offence decisions while ensuring ethical implementation in legal contexts.
• Conduct a structured engagement process, including workshops and empirical data collection from 300+ survey participants, integrating diverse insights into AI development.
• Establish a replicable model for public engagement with complex technological and ethical issues, contributing to policy discussions on AI governance in legal decision-making.
• Foster further funded research and collaborations on algorithmic justice, enabling expansion into more complex legal domains and international standards.
• Create a lasting impact on public discourse around AI ethics, adapting engagement methodologies, including art and behavioral science, for AI applications in other sensitive areas.

Partners and Collaborators

• The Dickson Poon School of Law
• Faculty of Mathematical and Natural Engineering Sciences
• Institute of Psychiatry, Psychology & Neuroscience
• King’s Digital Lab
• King’s Culture
• Science Gallery London
• Undergraduate and Master’s Students in Law and Mathematics
• Somerset House Studios, Artist Vivienne Griffin
• Transform Justice

Project Team

• Cari Hyde-Vaamonde, Project Lead
• Dr Pierpaolo Vivo, Project Co-lead
• Professor Dan Hunter, Project team member 
• Paul Caton (Digital Lab), Project team member 
• Geoffrey Noël (Digital Lab), Project team member 
• Rose Hepworth, Project team member 
• Jennifer Wong, Project team member 
• Mariia Bogdanova, Project team member 
• Penelope Gibbs (Director, Transform Justice), Project team member 

This project is supported by the One King’s Impact Fund

The One King's Impact Fund is part of One King’s Impact, King’s strategic programme to support and accelerate work within and beyond the University which creates positive change for people, planet and society.

AI in the justice system responds to the following Impact Priorities: 

Peace, Justice & Security  | Technology for Good