Judicial Guidance: Artificial Intelligence
On 12 December 2023 the Courts and Tribunals Judiciary published new guidance for Judicial Office Holders for the use of Artificial Intelligence (A.I.). Such a step may be perceived to show the Courts willingness to embrace the advances of A.I. but, at a closer glance, presents a cautionary approach.
The guidance is said to set out the key risks and issues associated with the use of A.I. and does so by highlighting areas of concern and the responsibilities of the user. By way of summary, the following are identified:
- Understanding A.I. and its applications
- Upholding confidentiality and privacy
- Ensuring accountability and accuracy
- Being aware of bias
- Maintaining security
- Taking responsibility
- Being aware that court/tribunal users may have used A.I. tools
These points are expanded upon in detail in the guidance to make clear exactly where things can go wrong and how to mitigate those risks.
For example, under confidentiality and privacy, it is warned that putting information into A.I. tools that are publicly available is equal to publishing information to the world, which raises concerns for confidentiality and data protection. This should obviously be avoided.
Other examples cited include tasks in which the Judiciary may find A.I. useful such as administrative work and summaries of large text. It also outlines where the use of A.I. would not be recommended such as in legal research and legal analysis. There are also telltale signs to look out for which can indicate that work has been produced using A.I such as unfamiliar citations, American spellings, overseas cases and substantive errors.
Ultimately, this guidance is intended to ensure consistency between using A.I. and the requirement for the Judiciary to maintain the integrity of the proper administration of justice.
As factual errors and inaccuracies are possible when using A.I. in the context of legal disputes, the emphasis remains on individual responsibility to verify material presented to and relied upon in Court. It is therefore easy to see why the Judiciary is cautious about the use of A.I. tools in resolving disputes.
