In the US, a personal injury lawyer significantly embarrassed himself by submitting documents to the federal court, citing no less than six cases that never actually existed—he had found these cases using ChatGPT. He had failed to check that these cases existed.
- The now-ubiquitous AI chatbot invented six legal cases.
- The lawyer, Steven A. Schwartz, said he’d never used ChatGPT and was unaware of its limitations.
- The danger of AI may not be that it develops a will of its own but that people double-check what the AI tells them.