At the end of February, I did an article, for a Coverage Opinions email blast, that shared the results of my ChatGPT insurance coverage test. I had asked the much talked-about artificial intelligence platform – that offers answers to questions in human-like conversation – what it knew about insurance coverage.
Very simple. I asked five very specific questions with answers that have been clearly stated by a state supreme court. The verdict – ChatGPT got 4 out of 5 wrong. But not just wrong – dead wrong.
I received loads of reader mail about the article. It was clear that the role of artificial intelligence in insurance coverage is on people’s minds. I decided to do a follow-up. It summarizes the first article – for those who didn’t read it – and then addresses the issue from a legal perspective and bad faith considerations.