The agent-assist popped up an answer that sounded so sure I sent it straight to the customer. Turned out it was flat wrong, and I spent the next day cleaning up the mess and apologizing. Now every time it suggests something, I hesitate, because I honestly can't tell when to trust it and when it's confidently lying.
More people experience this than they realize.
You rely on AI assist but can't tell when its confident answers are actually wrong.
If this sounds familiar, the Library can help you find the bigger picture.