Bring A Trailer Listing Got Burned By AI Photos, And It Won’t Be The Last Time
The digitally “enhanced” Cadillac sedan in the auction listing featured AI glitches, including one image that showed twin column gear shifters

by Chris Chilton
- BaT pulled a Cadillac auction after users flagged AI images.
- Photos showed cobblestone mats and a missing slam panel.
- Seller likely used AI for a cleaner, more polished backdrop.
The Bring a Trailer auction site is used to being at the center of enthusiast chatter, but this time, it’s not because it’s listing a car with an incredibly rare feature, but one with a feature that never existed until AI dreamed it up.
A 1999 Cadillac DeVille listing went live this past weekend, and it didn’t take long for the detail-obsessed enthusiasts who live in the comments section to notice that something was slightly odd about some of the pictures, and very odd about the rest of them.
Related: Just Eight AI Traffic Cameras Caught 29,000 Offenders. The Fines Go Straight To Your Phone
Major AI glitches that seem to not have been noticed by either the seller or Bring a Trailer’s upload team include floor mats that matched the stone driveway underneath the car, and an engine bay shot where the entire front slam panel is missing, as if the car was in a breakers’ yard. At least one image appeared to show two column shifters. Users started calling out the nonsense immediately.
“Come on guys, we all know the 1999 DeVille is 20 feet long on the left side, 14 feet long on the right side, somewhere between 25 and 30 pavers wide, and could be ordered with the rare A-pillar delete,” user @DavidRoseHRL wrote in the comments.
“BaT’s credibility is crashing harder than Richard Hammond,” wrote another.
Bring a Trailer Owns It
To BaT’s credit, staff did not pretend nothing happened. Their early response said they were “working with seller to obtain additional images,” which only poured fuel on the fire, because if the gallery is suspect, “more photos” is not a fix.
Not long after, BaT ended the listing, explaining that because the seller could not provide more photos quickly, it was best to stop the auction since it “cannot allow this auction to move forward in good faith.”
More interesting was BaT’s longer statement from Howard Swig, Head of Auctions. He called it a “huge error” with “multiple points of failure,” emphasized that BaT’s curation and quality control are “very human” with no reliance on algorithms or AI. “We screwed up here and will own that,” he said.
Not a Scam – This Time
There’s no suggestion that the seller or BaT wanted to deceive buyers. It looks like the seller just wanted a nicer, more aspirational backdrop for his images. Same car, more clicks, higher bids. In theory, AI could enhance the images, but instead, it introduced lots of errors and removed the trust in the world’s best-known enthusiast auction platform and what was probably an honest 81,000-mile (130,000 km) used Cadillac.
Online auctions depend on pictures as evidence of the condition and spec of the car, often telling us more than the written text. But when buyers start to doubt those images, they start to doubt the whole process.
“One of the core reasons many of us trust Bring a Trailer is the consistently high standard of listings and the expectation that sellers provide genuine, well-vetted content, @Tifosi wrote in the comments section below the listing.
“Allowing a listing of this nature to remain live undermines that trust and reflects poorly on the platform as a whole.”
Probably Not the First Time, Definitely Not the Last
BaT will almost certainly tighten policy and add screening, but this won’t be the last time AI slips into the gallery. We’re willing to bet that while this listing might be the first to be called out, the tech has probably already been used on different selling platforms across the world already without raising suspicions. As one of my Carscoops colleagues put it, it’s just a surprise it’s taken so long to go wrong in a big way.
Frankly, as the technology keeps evolving, it’s only a matter of time before we fall for it too here at Carscoops. Maybe we already have and just haven’t noticed.
We’re all still in the awkward early days of AI, trying to work out what to do with this new digital toolkit and how it can help us. But like any learning process, it’s not going to be smooth like fresh asphalt, it’s going to be bumpy like…cobblestones.
Take a look at the full picture gallery in the listing and see how many glitches you can spot.
![]()
![]()