A New York lawyer has been ordered to explain why he should not be disciplined after using ChatGPT for legal research.
The original case in question involved a man suing an airline over an alleged personal injury. His legal team, made up of lawyers from the firm Levidow, Levidow & Oberman submitted a court document on his behalf. The document contained a list of legal cases to prove, using precedent, why the case should be considered.
However upon closer inspection, the airline’s lawyers said they could not find a number of the cases mentioned in the brief.
“Six of the submitted cases appear to be bogus judicial decisions with bogus quotes and bogus internal citations,” Judge Castel wrote in an order demanding the man’s legal team explain itself.
More from News
- Growing Firms Slower To Fix Their Most Serious Software Flaws, Research Finds
- Can Anthropic Go It Alone? What Nvidia’s Crisis Means For Its Big IPO
- Shaking Up Ride Hailing: Zipo Is Officially Live In London
- Exabeam Advances Exabeam Sherpa Into AI-Guided Partner Operating System
- Has Microsoft’s Latest Outage Exposed How Dependent Businesses Have Become On Technology?
- COP31’s Antalya AI Pledge Asks Tech Companies To Disclose Their Energy Use – Is A Voluntary Promise Enough?
- Jensen Huang Is Out Amid Spy Allegations. How Bad Could The Fallout Get For Nvidia, AI And Global Chip Supply?
- Brits Want AI to Help Not Take Over New Research Finds
Despite ChatGPT having a disclaimer that it can produce inaccurate information, the lawyer who used the tool told the court he was “unaware that its content could be false”.
Mr Schwartz, a lawyer at the firm who prepared the document, said that he “greatly regrets” using the chatbot, which he had never used before.
He has vowed to never use AI to “supplement” his legal research in future “without absolute verification of its authenticity”.
The event comes as a warning that although AI like ChatGPT is a useful tool for information gathering, it should be used with caution, especially in professional environments.
Whilst it can provide human-like responses, it uses the internet as it was in 2021 as its database meaning some of its information can be outdated.
