How do I explain an AI risk to a team that isn't technical?
I'd explain AI risk in terms of the task your team does. Words like hallucination and bias can be familiar and still leave someone unsure what to do next. Describe the mistake in ordinary language and say who would be affected, then show where the result gets checked and where the work stops.
Start with a task the team recognizes. In a customer response, the problem might be quoting a policy that has expired. In a report, it might be an estimate presented as a measured result.
A recommendation can leave out a relevant group of people. Name where your team can check the result and who the reviewer is, then mark the point where work must stop if a fact is unresolved.
Use a realistic example in training and let employees find the issue and explain their next step. Dramatic examples can make ordinary mistakes seem too minor to report, so skip them.
Keep the technical vocabulary around for specialists who investigate. The everyday guide should let a person recognize the problem without mastering an AI glossary.
What to try this week
Imagine the AI mistake that could happen in one task your team does, then describe it in ordinary words. Name who would be affected, where it gets checked and when work must stop.