Judges given green light to use ChatGPT in legal rulings
Judges in England and Wales have been given official guidance allowing them to use ChatGPT, an artificial intelligence language model, to assist in writing legal rulings. Despite concerns that AI can generate fictitious cases, the Judicial Office believes that ChatGPT can be useful for summarizing large amounts of text and for administrative tasks. However, cautionary measures have been outlined, as chatbots like ChatGPT are deemed inefficient for research and are prone to inventing bogus cases or legal texts. The guidance also highlights potential misuse of bots by the public when initiating legal cases, as well as the risk of deepfake technology being employed to fabricate evidence.
The Master of the Rolls, Sir Geoffrey Vos, acknowledges the significant opportunities that AI presents in developing a more efficient and cost-effective justice system. He emphasizes the need for judges to be aware that while AI can provide accurate responses, it can also yield inaccurate ones. Lord Justice Birss, a senior judge, has personally found ChatGPT jolly useful, using it to summarize an area of law he was already familiar with and incorporating it into a court ruling. However, he assures that no confidential information was entered into the chatbot.
Concerns about chatbots producing fictional cases have arisen from instances in the United States where lawyers have encountered difficulties after utilizing ChatGPT to draft court filings. As AI language models generate sentences based on vast amounts of text they have been trained on, there is a risk of them inventing facts. The legal guidance explicitly states that current public AI chatbots lack convincing analysis or reasoning capabilities. Judicial officials are individually responsible for the material attributed to them.
The guidance warns against inputting private information into chatbots, reminding judges that it could be made public. It also acknowledges that chatbots often provide answers based on US law when asked legal questions, and the presence of American spelling or references to US cases may indicate an AI-generated response. In terms of responsibility, Sir Geoffrey underscores that lawyers utilizing chatbots to produce false evidence may still face perjury charges or criminal sanctions.
The use of AI in legal rulings raises both opportunities and challenges. While AI can expedite and streamline processes, the potential for inaccuracies, fictional cases, and fabricated evidence must be addressed. Public understanding of the limitations of AI and vigilance in its application are crucial. As technology continues to advance, the judiciary must remain knowledgeable and adapt to ensure a fair and just legal system.
In conclusion, the Judicial Office’s guidance permits judges in England and Wales to utilize ChatGPT in their legal rulings. However, caution is warranted as the technology has been found to create fictitious cases. While AI offers opportunities for a better digital justice system, it is essential that judges and lawyers exercise discretion in using AI as a tool, recognizing both its benefits and limitations. By staying informed and embracing responsible AI usage, the legal system can leverage technology to enhance efficiency and effectiveness while maintaining the integrity of justice.
