AI Can Make Mistakes: Stories from a Test Engineer

Everyone talks about how AI can make developers and testers more productive. Fewer people talk about what happens when AI is wrong. Over the last year, I started using AI for everything from generating API tests and automation code to debugging failures, reviewing requirements, and even helping me bake cookies. Sometimes it saved hours of work. Sometimes it confidently suggested APIs that didn't exist, generated tests that looked perfect but missed critical defects, misunderstood requirements, or recommended recipe changes that turned out to be a terrible idea. In this session, I'll share real examples of where AI helped, where it failed, and what those failures taught me about using AI responsibly in software engineering.