Separating parents who turn to tools like ChatGPT for legal advice have no route to effective redress if the information they are given is wrong, according to a new report by Dr Natalie Byrom. The report makes urgent recommendations for government to address the risks of digital tools head-on, in order to both strengthen protections for families and the justice system and to harness the benefits AI could bring.
Today Nuffield Family Justice Observatory convened a discussion on the use of AI and digital tools in the family justice system with MP Jake Richards, Minister for Sentencing, Human Rights and Technology for the Ministry of Justice; Dr Natalie Byrom, an independent researcher and policy adviser whose work focuses on how data and technology are reshaping access to justice; Lisa Harker, Director of Nuffield Family Justice Observatory; Olive Craig from Rights of Women, and Gaia Marcus, Director of the Ada Lovelace Institute.
Government and the court service are increasingly reliant on technology that is built, owned and managed by the private sector – a significant departure from previous approaches to digitising the justice system, which championed publicly built, publicly owned technology as an active design choice. There is concern that regulation and governance is failing to keep pace, resulting in inadequate protection for families or the justice system.
Delegated by default? Privately owned technology in the family justice system by Dr Natalie Byrom and published today by Nuffield Family Justice Observatory warns that in family justice this ‘delegation by default’ approach creates risks that are notably acute due to both the nature of disputes and the vulnerability of the people coming to court.
The growing use of GenAI tools like ChatGPT and Claude for search, risks disrupting decades of investment aimed at improving access to authoritative, accessible dispute resolution advice online. In family justice as elsewhere, there is emerging evidence that litigants are turning to these tools directly for advice and document preparation, despite the fact that GenAI tools frequently get the law wrong. These tools are also prone to affirming users’ beliefs even where they are incorrect, often discouraging behavioural repair and undermining ongoing attempts by policymakers and professionals to de-escalate conflict and divert parents towards conciliation.
The report is clear that the risks aren’t an argument against digital or AI-mediated justice, or against a role for the private sector in delivering it, but that a deliberate approach is needed which manages the opportunities and risks.
