The introduction of ChatGPT in November 2022 has taken the everyday use of artificial intelligence (AI) to a new level and with new technology comes new professional indemnity risks – just ask law firm Levidow, Levidow & Oberman, who have made the news for all of the wrong reasons. In an unprecedented and highly publicised case, US attorneys, Steven Schwartz and Peter LoDuca and their firm were asked by US District Court Judge to show cause why they should not be sanctioned for their reliance on six cases generated by ‘research’ undertaken by ChatGPT. The cases simply did not exist.
Alarmingly LoDuca also signed an affidavit deposing to the authenticity of the case authority, relying on Schwartz’s research via ChatGPT.
One of the (now) well known risks of using ChatGPT is that it ‘hallucinates’; that is, it makes up facts and sources.
According to the attorneys, they did not consider the possibility that an AI tool like ChatGPT could provide false information. The convincing information from ChatGPT included purported excerpts from fictitious cases. The law firm sought to verify the existence of the cases – from ChatGPT itself – which is a little like asking a known fraudster whether the company funds are safe with them.
The US example is startling in a profession which is traditionally risk averse. However, it is really a new example of an age-old PI risk. You should not rely on case law authority and submissions produced by junior lawyers without checking the source material, so why would you rely on authority produced by a Chat bot before providing advice to clients or relying on the cases in Court?
False information is not the only risk of emerging AI technology, with the data regulator in Italy banning ChatGPT due to privacy concerns.
Until the limitations and risks of AI are properly understood, law firms are well advised to avoid reliance on AI technology in providing legal advice and other legal services. Not only may lawyers face negligence claims and disciplinary action but the worldwide infamy of the Levidow case has shown the massive damage a firm’s reputation can sustain.
For more information on the risks of AI in the legal profession, please attend the presentation by Justine Siavelis as part of the Risk Management Training Programme 2023/24 at Law Mutual (WA). (Booking details available soon.)
|