Should you trust AI with your divorce?

Whilst this is a huge subject for discussion and debate, we cannot ignore that AI is making its way into our homes, offices and now our court rooms daily.

We are regularly seeing people asking AI for advice, using AI to assist with the completion of forms and letter writing, or relying on advice from AI in court, especially when they are not represented by a solicitor. 

In this article, we are going to look at some of the practical risks for those relying solely on AI, especially in court proceedings and when negotiating settlements.

Large language models (like ChatGPT) can produce text that sounds authoritative but is factually incorrect, also known as an ‘hallucination’ One widely reported example involved a US lawyer submitting a legal brief that relied on a fictitious case citation generated by ChatGPT, which resulted in professional sanctions. We are also regularly encountering individuals attempting to rely on incorrect or misleading information obtained from AI tools.

AI can easily cross the line from ‘information’ to legal advice.  Someone relying solely on AI might end up filing the wrong documents, misunderstanding a document, or missing key elements of an active case. In addition, it might also result in a solicitor, Judge or other professional thinking that someone has a higher level of understanding than is true.  That itself can be worrying, as it is important that both parties fully grasp the meaning and potential repercussions of any agreement they are entering.

Another potential risk arises where AI generates information based on the wrong jurisdiction. Family law varies significantly between countries, and AI tools may provide responses that appear accurate and persuasive but are based on foreign legal systems or outdated rules. This can lead to parties relying on procedures, legal principles or entitlements that simply do not apply to their case.

Another issue relates to privacy and confidentiality. If someone uploads data to a public AI tool, that is not private. In most situations, it will be a breach of court order to share the data with a third party. From a professional perspective, it would also raise concerns around GDPR breaches. Once uploaded, the data can be stored, shared and used for training purposes.

There is little doubt that AI will continue to develop and integrate into all areas of law. However, from a practical perspective, caution should be exercised. Those choosing to use AI should remain vigilant, carefully fact-check information, and seek appropriate professional advice.