YONKERS, NY – A viral social media prank using AI-generated images of fake “homeless men” and “intruders” has triggered real-life panic in Yonkers, prompting swift warnings from the Yonkers Police Department.
Officers say people are using artificial intelligence to create realistic photos of strangers standing in doorways, sitting on couches, or lurking inside homes — then sending them to unsuspecting roommates or relatives as a prank.
Some of those recipients, believing they were witnessing a real intruder, have dialed 911, leading to urgent, lights-and-sirens police responses.
By the time officers arrive, they discover the supposed trespasser exists only as pixels on a screen.
Police say pranks waste resources and risk lives
According to the department’s public safety alert posted Friday, officers in Yonkers have already responded to several false emergencies tied to the trend. Officials warned that the calls not only waste valuable resources but also create dangerous situations for both residents and responding officers.
“These are not harmless jokes,” the statement read. “When our officers respond to a call of an intruder, they are acting on the belief that someone’s safety is at risk. That split second of misinformation could end tragically.”
A national warning on AI misuse
The so-called ‘AI homeless man’ prank appears to be spreading beyond New York, with other departments across the country reportedly seeing similar cases. Experts have cautioned that while AI tools can be entertaining and educational, they can also create highly convincing fabrications that blur the line between real and fake emergencies.
Police are urging parents to talk to their children about using AI responsibly, stressing that intentionally creating a false emergency can lead to criminal charges.
Officials urge public to verify before dialing 911
Authorities recommend pausing before calling 911 if an alarming photo or message surfaces — and verifying its authenticity. The department suggested reaching out to family members or roommates first to ensure the situation isn’t part of a prank.
“Talk to your kids,” the department’s post concluded. “AI can do amazing things — but misusing it to fake an intruder crosses a dangerous line.”