Receive Free Artificial Intelligence Updates
We will send you a myFT Daily Summary email summarizing the latest Artificial intelligence news every morning.
The author sits on the Civil Justice Council and the Senior Data Governance Committee of the UK Ministry of Justice
ChatGPT, the technology that launched a thousand dilemmas, has now caught the attention of the UK’s most experienced judges. This month, the House of Lords Constitution Committee heard testimony from the Chief Justice. He described the experience of a litigant in Manchester who relied on ChatGPT for legal advice and inadvertently made an entirely fictitious submission to court. Similar experiences have been reported in the United States.
Concerns about generative AI models returning accurate legal information have led even ardent proponents of the technology to conclude that it should only be used to augment, rather than replace, advice from lawyers. This is a pragmatic approach, even if it undermines its potential to expand access to legal advice to the growing number of people who cannot afford it. However, in the rush for this solution, justice system officials are missing an opportunity to question and address Why the tool returns inaccurate advice in the first place.
The root of the problem stems from the model’s inability to access enough relevant legal information. Legal tech experts are clear that training generative AI on relevant documents, such as judgments and court decisions, can significantly improve the accuracy of its answers. However, in England and Wales, access to these judgments and decisions is difficult to obtain. The most comprehensive collections are owned by commercial publishers who charge high access fees. Content held behind a paywall far exceeds that available to the public – a 2022 study reported that only half of judicial review judgments are available on BAILII, a free website.
While the government has now invested in a service to make more information available, the decision to only publish cases that judges consider important and want online has led to problems. A study published in January 2023 found that only three-quarters of judgments issued between May and July 2022 were available on the new open-access website. Unless urgent action is taken, the asymmetry between the information available to those who can pay and that available to those who cannot is likely to persist indefinitely.
The lack of public access to a full record of court judgments and decisions has long been of concern to journalists and transparency advocates. The era of generative AI creates a new imperative to tackle the problem. If we don’t, the best, most accurate and powerful AI tools will be developed by and for those who already have access to legal information – private publishers, insurance companies and mass litigants. This will exacerbate and entrench existing inequalities.
Anyone doubting the benefit of pre-existing access to legal data need only look to the recent acquisition of Casetext by Thomson Reuters. In June, Thomson Reuters acquired this relatively small company, which only recently specialized in generative AI. On a call with shareholders, the executives justified the size of the deal by saying that “they are one of the few corporate groups that also have access to data.” They clearly felt that the head start offered to Casetext by its ability to train ChatGPT4 on legal data justified the $650 million price tag.
Lawyers, judges and even the occasional justice minister like to quote clause 40 of Magna Carta – “to no one we will sell, to no one we will withhold or delay right or justice”. But while justice may not be for sale, access to court-generated information certainly is. In the age of AI, access to justice and access to legal information are increasingly intertwined. If we are serious about harnessing new technologies to address inequities in our legal system, we need to start by determining who can see and use legal data.
