Seven questions to ask before you trust an AI answer about a house.

NAR’s own guidance on AI home search says recommendations should be treated as a starting point for research, not a substitute for professional expertise, and that fair housing laws, MLS rules and data accuracy still apply. Here’s a practical version of that.

1. 1Where did that number come from? An MLS record, a county record and a computer estimate are three different things. If the tool doesn’t say which, ask.
2. 2What date is it? Markets move. The 30-year rate went from 6.34% to 7.28% in a year. See the October rate check.
3. 3Is the photo altered? In California, it should be labeled. See how to read a listing photo.
4. 4Is the home actually available? A pre-market listing can be visible but not tourable. See the Preview post.
5. 5Who benefits from this answer? A listing site’s chatbot is a front door to its own listings. See the ChatGPT apps post.
6. 6Does it describe places, or people? A good answer talks about commute, price, parks and schools. If a tool starts describing who lives in a neighborhood, stop. That’s a fair-housing problem, and any professional using AI is responsible for avoiding it.
7. 7Can a person check it? In Cotality’s survey, 44% of buyers said they’d pay extra to have a human verify what AI told them. It’s a reasonable instinct.