This is how to use AI before buying a used car
Two listings, one budget, and no way to tell which cheap car gets expensive later. Compare ownership-cost categories before you even call.
This is how to use AI before buying a used car.
Two tabs open. A 2014 sedan with 128,000 miles for $8,900, and a 2016 with 96,000 for $9,400 that mentions "minor cosmetic damage" and nothing else. The prices are twenty minutes apart; the next five years of ownership might not be. The listing that looks cheaper is sometimes the expensive one wearing a good price.
Paste the public facts from both listings, plus your budget and how you drive, and make the comparison about categories instead of stickers:
Copy this prompt:
I’m comparing two used cars with a $10,000 ceiling and a 40-minute daily commute.
Car A: 2014 sedan, 128,000 miles, $8,900, two owners, "recent brakes," no service records mentioned.
Car B: 2016 sedan, 96,000 miles, $9,400, one owner, "minor cosmetic damage," no photos of the damage.
Do not pick a winner. Instead: 1) the ownership-cost categories I should compare for cars of these ages and mileages, 2) what each listing is NOT saying that I should ask about, 3) the questions to send each seller before I spend anything on an inspection, and 4) what a pre-purchase inspection should focus on for each car specifically.
Swap the details for your own listings. The structure is the part that matters.
Level up the ask:
One more move before you send it. Ask this first:
Rewrite my prompt above into a version that will get me a better result. Tell me what you changed and why, then answer the improved version.
What a useful answer reveals
A good reply does not crown Car B. It maps the decision: which model years carry known issues worth asking about, which maintenance items come due near each car’s mileage (and roughly what they cost), and, most usefully, what each listing’s silence means. "Recent brakes" with no records is a question, not a feature. Unphotographed "cosmetic damage" is a question with a dollar sign on it.
Every gap becomes a message you can send before driving anywhere. Sellers who answer specific questions plainly are telling you something; so are sellers who don’t.
Turn it into the seller messages
Copy this prompt:
Turn those gaps into two short, polite messages, one per seller, each with numbered questions I can send as-is. Ask for service records, the story behind the flagged items, and whether they’re open to a pre-purchase inspection at my mechanic.
A seller’s answer to the inspection question is the most informative sentence in the whole exchange.
Where AI helps, and where it hands off
AI is strong at categories, known patterns for a model year, and questions you would not think to ask. It cannot see rust, hear a lifter tick, or know whether this specific car was loved. That is the pre-purchase inspection’s job, and at roughly $100 to $200 it is the best money in the entire process. Anything the chat concluded is preparation for that appointment, not a substitute for it.
- Share only: Public listing facts and photos, your rough budget, and how you actually drive; never the VIN, the seller’s contact details, private messages, or financing paperwork.
- Verify: Real condition with a mechanic’s pre-purchase inspection and an official vehicle history report before any money moves; price against sold comparables, not asking prices. The inspection, not the chat, decides.
This is how to use AI.