With the huge advancements in artificial intelligence (AI) over recent years, most notably with the release of ChatGPT (a new and improved ChatGPT-4 was released to paid subscription users on 13 March 2023) many are wondering how it will affect the legal sphere.
Undoubtedly, the legal industry must be ready to harness the advantages that it could offer, although there are clear limits to the ‘service’ AI can currently provide.
Putting ChatGPT to the test
At present, if you ask ChatGPT for advice in relation to fact-specific circumstances of an example divorce scenario, it will “recommend that you seek legal advice from a family law solicitor who can provide you with specific advice based on your individual circumstances”.
When pressed in relation to specific outcomes, the bot is clear that it “cannot predict the outcome of any legal proceedings” and that independent advice from a suitably qualified lawyer is consistently encouraged “to advise [you] on the best course of action to take to achieve the outcome [you] desire”.
One of the main problems with any new advancement is that there will inevitably be numerous issues with the necessary regulation of the developing technology. In family law, confidentiality is of particular concern. Financial disclosure is of central importance during a divorce.
While solicitors and barristers are under strict professional obligations (lawyer-client confidentiality being sacrosanct), how could this apply to ChatGPT or other AI models offering basic legal advice and how would this be regulated in practice? And without being able to draw on specific details, it is impossible for any form of AI to provide tailored legal advice. Similarly, without exposure to different issues or data sets, the AI language model will not be able to adapt and improve its responses.
