No. of Recommendations: 6
had begun to use artificial intelligence to make the fraud referrals to him
It's funny. Just last week in the sauna at the gym, I had a conversation about "AI" with a guy that works in insurance claims. He also said they are using "AI" now, but as the conversation progressed, he said that the "AI" apparently determined that the best indicator of potential fraud was a very simple determination - basically it is "has the claimant made a claim in the last 5 years?". If the answer is yes, it's more worth investigating, if the answer is no, it is less worth investigating. So, the core of that AI is rather trivial for a first-order approximation.
