A night at the casino turns into a nightmare

What began as a routine night of gambling for 40‑year‑old UPS driver Jason Killinger at Reno’s Peppermill Casino turned into a wrongful arrest after the casino’s facial recognition system labeled him a trespassed patron named Michael. According to body‑camera footage and police records, Peppermill security detained Jason after their artificial intelligence software returned a “100% match” between his face and a previously banned customer.​
Despite Killinger handing over a valid Nevada Real ID driver’s license, security insisted he was the trespassed man and called Reno Police to have him arrested for criminal trespass.​

“The software is pretty cool”: why officers trusted the AI

When Officer Richard Jagger arrived, he was told that Peppermill’s “pretty sophisticated” AI facial recognition system had positively matched Jason to Michael, a man barred months earlier for sleeping on casino property.​
The officer repeatedly referenced the technology in his conversations, saying the system had produced a 99% and then a 100% match from different camera angles, leading him to conclude that “a reasonable and prudent person” would believe Jason and Michael were the same individual despite discrepancies in their identification.​

Those discrepancies were not minor. Jason’s Real ID listed him as seven years younger, two inches taller in one comparison, and 50 pounds heavier than Michael, with different eye colors and signatures.​
Still, Jagger and casino staff focused on the AI result and the apparent similarity of the photos, speculating that Jason might have a fake ID, a “hookup at the DMV,” or two legitimate licenses under different names.​

Detained, cuffed, and disbelieved

Jason consistently insisted that he was not Michael and that he had been a regular craps player at the Peppermill for years without incident.​
He told the officer he drove semis for UPS, produced his Nevada driver’s license, and begged to be taken out of handcuffs because his shoulders hurt, only to be told he would “sit there for a minute” while the investigation continued.​

When a records check confirmed that both Jason’s and Michael’s licenses were valid DMV entries, Jagger’s doubts should have deepened—but instead he treated the situation as further proof of fraud.​
Even as Jason pointed out that he had a paycheck stub, vehicle registration, union card, DOT medical card, and insurance documents in his own name, the officer dismissed these as insufficient to counter the AI‑driven suspicion.​

Ultimately, Jagger concluded he “could not identify” Jason, despite the Real ID and multiple corroborating documents, and arrested him as a John Doe so the jail could “figure it out” with fingerprints.​
On the ride to jail, the officer admitted on camera that he “kind of believe[d] him” but felt constrained by the casino’s system and his own belief in the technology’s reliability, noting that similar systems had been used to catch serious offenders in the past.​

Fingerprints prove the AI wrong

At the jail, Jason’s fingerprints quickly confirmed what he had been saying all along: he was, in fact, Jason Killinger, not the previously trespassed Michael.​
He was released later that day on his own recognizance, but only after hours in custody and a trip to urgent care, where he was diagnosed with strains to both shoulders and contusions on his wrists from the restraint.​

Despite the confirmed misidentification, the paperwork on his arrest was changed from “John Doe” to “Jason Killinger,” meaning prosecutors treated him as if he were the original trespasser.​
Although the trespass charge was eventually dismissed, Jason was left with a criminal record connected to an incident he did not cause, and with injuries and emotional trauma from an arrest that never should have happened.​

Lawsuit and wider concerns about facial recognition

In July 2025, Jason filed a federal civil‑rights complaint against Officer Jagger, arguing that the arrest violated his rights by relying almost entirely on flawed facial recognition and discounting clear, independent evidence of his identity.​
The complaint also notes a broader pattern: wrongful arrests linked to facial recognition systems have already been documented in states including Louisiana, New Jersey, Maryland, and Texas, where misidentifications led to innocent people being jailed.​

Civil‑liberties advocates and some city governments now require that facial recognition matches be treated only as investigative leads, not as stand‑alone probable cause. Many jurisdictions insist officers obtain separate, non‑AI evidence—such as corroborating witness statements, physical evidence, or verified identity documents—before making an arrest.​​
Jason’s case illustrates what can happen when that safeguard is ignored: sophisticated software and human bias reinforce one another, and an innocent person becomes easier to arrest than to believe.

Jason Killinger’s federal civil‑rights case against Reno police officer Richard Jager is still active and moving through pre‑trial stages, with discovery underway and a trial expected but not yet held.​

Case posture and recent filings

• Killinger filed his lawsuit in U.S. District Court for the District of Nevada in mid‑2025, alleging constitutional violations tied to his wrongful arrest at the Peppermill Casino.​

• Court docket entries show the parties have submitted a joint discovery and scheduling plan, and a case‑management conference that was previously canceled has been reinstated by court order, which signals the case is proceeding rather than being dismissed or settled.​

Parties, claims, and settlements so far

• Killinger has already settled separately with Peppermill Casino, reportedly for an undisclosed amount, so the remaining active defendant in this lawsuit is Officer Jager (and, effectively, the City through him).​

• The complaint accuses Jager of fabricating and omitting key facts in his reports—such as downplaying or leaving out the fingerprint results that cleared Killinger—and claims these actions violated Killinger’s Fourteenth Amendment due‑process rights and support claims for compensatory and punitive damages.​

What to expect next

• Reporting on the case notes that, barring a last‑minute settlement, the suit is expected to go to trial sometime next year, after the close of discovery and any summary‑judgment motions.​

• Until the court issues further orders or a verdict, Killinger remains in the posture of a plaintiff actively pursuing damages for wrongful arrest, alleged report falsification, and the lingering criminal record created by the misidentification.​

Lessons from the Peppermill case

The Peppermill incident underscores several key lessons about AI in policing and private security:

• AI is not infallible: A “100% match” can still be wrong, especially in environments with variable lighting, camera angles, or look‑alike subjects.​

• Human judgment must come first: Officers had access to a Real ID license and multiple corroborating documents, yet chose to trust software over tangible evidence and the suspect’s consistent story.​

• Policies need clear limits on AI: Without rules requiring independent corroboration, facial recognition can become a shortcut that substitutes probability for proof and convenience for constitutional safeguards.​​

Jason Killinger’s ongoing legal battle will help determine how courts view the use of facial recognition in probable cause decisions—and whether future officers can justify an arrest on the strength of an algorithm that “feels” right, even when the facts say otherwise.