TL;DR: The Heber City Police Department in Utah started using AI software to write police reports from body camera footage. In January 2026, one report claimed an officer had "shapeshifted into a frog." The AI had picked up background audio from Disney's "The Princess and the Frog" playing on a TV. The department says the tool saves "six to eight hours weekly." Experts say the real danger isn't frog hallucinations. It's bias amplification, reduced accountability, and the impossibility of auditing AI-generated reports.
How an Officer Became a Frog
In December 2025, the Heber City Police Department started testing Draft One, an AI tool made by Axon that uses OpenAI's GPT models to generate police reports from body camera audio [1].
In early January, something went wrong. An officer responded to a call where Disney's "The Princess and the Frog" was playing in the background. The AI didn't understand that the movie dialogue wasn't part of the police encounter.
The result: an official police report stating the officer had "morphed into a frog" while conducting police activity [2].
"The body cam software and the AI report writing software picked up on the movie that was playing in the background, which happened to be 'The Princess and the Frog,'" Sergeant Rick Keel explained to reporters [3].
The department issued a correction. They don't employ amphibious officers.
Why This Isn't Actually Funny
A frog transformation is easy to catch. But what about hallucinations that sound plausible?
American University law professor Andrew Ferguson told the Associated Press he's concerned that "automation and the ease of the technology would cause police officers to be sort of less careful with their writing" [4].
The real danger isn't comedy errors. It's:
- Subtle inaccuracies that slip through review
- Bias amplification from training data
- Reduced accountability when officers can blame AI
- Impossible auditing of AI-generated content
If AI adds a word like "aggressive" or "threatening" to describe a suspect's behavior, and an officer signs off without careful review, that language becomes part of the permanent record. It influences charging decisions, plea bargains, and trials.
The Company Behind the Frog
Draft One is made by Axon, the same company that makes Tasers and body cameras for police departments across America [1].
Aurelius Francisco, co-founder of the Foundation for Liberating Minds, finds this troubling: "The fact that the technology is being used by the same company that provides Tasers to the department is alarming enough" [5].
Axon has an obvious incentive: sell departments body cameras, then sell them AI software that processes body camera footage. It's a vertical integration of law enforcement technology.
The Electronic Frontier Foundation reviewed Draft One and found the software "seems deliberately designed to avoid audits that could provide any accountability to the public" [4].
AI in Policing Is Spreading Fast
Heber City isn't alone. Law enforcement agencies across America are adopting AI for:
- Report writing from body camera footage
- Facial recognition to identify suspects
- Predictive policing to target neighborhoods
- "Behavioral detection" to flag suspicious activity
- Weapon scanning in transit systems
The New York MTA just issued a request for AI to monitor 15,000 subway cameras. The pattern is clear: AI is becoming standard police infrastructure, often with minimal oversight.
The "Time Savings" Trade-off
Despite the frog incident, Sergeant Keel says Draft One saves him "six to eight hours weekly now" [2].
That's the trade-off police departments see: faster reports mean officers spend more time on patrol.
But faster isn't always better. A rushed, AI-generated report that an officer skims and approves could contain errors that ruin someone's life. A wrongful arrest based on AI-hallucinated details. A trial influenced by machine-generated characterizations of events.
The Heber City trial run ends soon, and officials expect to keep some form of AI reporting. They haven't decided which system to use [3].
What You Can Do
Request Police Reports
If you're ever involved in a police encounter, request a copy of any report. Check it carefully for accuracy. Errors can be corrected early. They're harder to fix later.
Know Your Local Policies
Does your local police department use AI for reports? Many don't publicize this. Ask your city council or submit a records request.
Support Transparency Requirements
Advocate for laws requiring disclosure when AI is used in police reports. If a document was generated by AI, defendants should know.
Document Your Encounters
If it's legal in your jurisdiction, record police encounters. Having your own evidence provides a check against AI-generated narratives.
The Bottom Line
An AI turning a cop into a frog makes headlines because it's absurd. But the mundane errors (the subtle biases, the plausible hallucinations, the unchecked assumptions) are what should worry you.
Police reports aren't just paperwork. They're the official record of what happened. They influence who gets charged, who goes to jail, and who's believed in court.
Handing that power to AI without strong auditing isn't just risky. It's reckless.
References
- Futurism - Cops Forced to Explain Why AI Generated Police Report Claimed Officer Transformed Into Frog (January 2026)
- UPI - AI-generated police report states Utah officer was turned into a frog (January 2026)
- Axios Salt Lake City - How AI turned a Utah police officer into a frog (January 2026)
- Above the Law - Police Wonder If AI Bodycam Reports Are Accurate (January 2026)
- Vice - An AI-Generated Police Report Claimed a Cop Transformed Into a Frog (January 2026)