Something is haunting your workshop.
It looks exactly like real information.
You're facilitating a session on workshop dynamics. Two learners are presenting their research. One of their slides has something hiding in it — confident, polished, and not entirely real. Shine your light carefully. Everything is not what it seems.
A professional development session is underway. Two learners are presenting research. Something in this room may not be what it seems.
AI hallucinations are outputs that sound completely real — specific, confident, well-formatted — but have no factual basis. They don't announce themselves. They don't flicker or glitch. They walk right through the wall like they belong there, wearing the clothes of credible information.
AI-assisted work has become a normal part of how people research, write, prepare, and present. For many educators and L&D professionals, this is actually how AI first shows up in their world — not as a new platform to learn or a policy to navigate, but as content. Someone else's content, already prepared, already in front of an audience. AI fluency begins with recognizing that moment for what it is. What matters then is what happens next — whether the output gets a second look before it gets passed along. Hallucinations thrive in that gap between what the AI said and what anyone stopped to verify.
Your organization is running a PD session on personality frameworks and participant dynamics in workshop settings — specifically, how to manage difficult moments when learners disengage or push back. Two learners, Lynn and Russ, have each prepared a short research presentation.
Both used AI in their process. That's where the story diverges. One leaned on it as a source. The other used it as a starting point and went further. As the facilitator, you won't be told which is which — your job is to read the room, evaluate what's on the slide, and respond.
You're not grading their presentations. You're practicing the habit of pausing before you accept what confident-sounding information puts in front of you — because in a live session, what gets through shapes what your learners walk away believing.
Notice what just happened. The slide looked professional. The claim was specific. Some heads nodded — and someone raised an eyebrow.
That eyebrow belongs to your room now. Whatever Lynn put on that screen, it lands in your session. How you respond — or don't — is what your participants will take with them. That's the weight of facilitation in an AI-assisted world: you don't always know where the content came from, but you're still the one standing at the front of the room.
One more presenter is waiting. Fair warning — not everything in this mansion is a ghost.
Two learners. Both used AI. Both mentioned it casually. Only one treated it as the source.
Lynn let a confident answer become a fact — her own experience felt like confirmation, and the AI felt like proof. Two feelings, zero verification. Russ used AI the way you'd use a knowledgeable colleague: it pointed him somewhere, and he went there himself.
The skill you practiced here was never "spot the lie." Sometimes the slide is clean. The skill is the pause — the willingness to ask "how do you know?" of any claim, any tool, and yourself. Especially when it already tracks with your experience.