Slower, costlier and less reliable: how to stop letting AI guess answers you already know

Would you trust a chatbot to tell you your refund rights? On the day his grandmother died, a man asked Air Canada’s chatbot about bereavement fares. It described a policy that didn’t exist, and the airline then argued it wasn’t responsible for what its own chatbot had said.

But the right answer was written down all along. So why do so many businesses let AI guess it, and pay for that in mistakes, bigger bills and slower replies?

In this week’s Signals from the AI Edge, I share what happened when I put that to the test, and how to get answers that are right every time, faster and for a fraction of the cost.