The Westbrook Police Department (WPD) in Maine had to apologize for using generative AI to alter a photo of items seized in a drug bust. They first denied the use of AI when called out, then later deleted the original post and their initial response and admitted to using ChatGPT.
The incident brought up fears of what potential future AI-assisted policing could look like. This fumble isn’t exactly inspiring confidence in cops’ ability to handle the technology.
Westbrook, Maine police department busted for using AI
According to the local ABC affiliate, the WPD posted the altered photo to their Facebook page on Sunday, June 29. On the following Tuesday, they published a follow-up message responding to multiple calls they received accusing them of creating the image with AI.
Some of the packaging on what the department displayed was blurry and in another language, making it look to some like the kind of gibberish ChatGPT tends to come up with when tasked with generating writing. The lettering on other items, however, could not be accounted for.
The department initially denied altering the image at all.
“We want to set the record straight — this is NOT an AI-generated photo,” they said. “What you see is what was seized by officers during the course of the drug arrests. Is the packaging weird and look like gibberish or a different language? Yes. But that is legitimately what was located and seized by officers.”
“We assure you that Westbrook PD is not and would never generate an AI photo to try and depict evidence that was seized.”
Admission and apology
While the entire photo may not have been dreamed up by AI, the WPD later deleted this response along with the original post and issued an apology. The follow-up admitted than an officer had plugged the real photo into ChatGPT in order to add the department’s badge on the wall.
They then claimed that, somehow, ChatGPT made a bunch of other changes on its own.
“Unbeknownst to anyone, when the app added the patch, it altered the packaging and some of the other attributes on the photograph. None of us caught it or realized it,” they claimed. “We apologize for this oversight. It was never our intent to alter the image of the evidence.”
The AI not only changed the lettering on some packaging, it removed stickers, a rubber band, and altered the appearance of every other item in the photo. This included lettering clearly written in English in the original.
As the story spread to X, those already critical of either police or AI jumped on it.
Popular user @CantEverDie warned that “cops already lie but now they’ll be able to completely fabricate evidence,” adding, “we need to heavily regulate generative ai.”
“I feel like even the most optimistic AI enthusiasts could agree that banning the police from using generative AI might be a good idea if we want to prevent some really dodgy sh*t from happening,” said @KindHeartedHunk.
The internet is chaotic—but we’ll break it down for you in one daily email. Sign up for the Daily Dot’s web_crawlr newsletter here to get the best (and worst) of the internet straight into your inbox.
