A 23-year-old Nebraska man has been sentenced to more than 10 years in federal prison after robbing an Omaha bank at gunpoint, with investigators later finding that he had used ChatGPT to research the crime, including police response times and how to handle the weapon.
The robbery was over in minutes. The evidence lasted longer.
On February 10, Lewis-Payne entered an i3Bank branch on Spaulding Street in Omaha carrying a black semi-automatic handgun.
He threatened two employees with the weapon and demanded money, warning them against putting dye packs among the cash. The tellers complied and he left with about $9,175.
But the robbery had already produced evidence that did not depend on identifying him through traditional police work.
Bank surveillance recorded his face because he was not wearing a mask. Cameras near the bank captured the vehicle he allegedly used, including its registration plate and unusual characteristics such as a toolbox and particular rust spots.
The investigation then followed the money.
After the robbery, surveillance footage and transaction records showed Lewis-Payne making purchases at Westroads Mall with cash. Investigators later found him near one of his residences, where he fled on foot and left behind banded cash.
A subsequent search uncovered the vehicle, additional currency and clothing investigators said matched what he had worn during the robbery, including distinctive red Nike slides.
Only $2,397 of the stolen $9,175 was ultimately recovered. The handgun was not found.
Then investigators looked at his phone
Lewis-Payne admitted to driving the vehicle but denied involvement in the robbery, according to federal prosecutors.
He consented to a search of his phone.
That search gave investigators another line of evidence: conversations showing that Lewis-Payne had used ChatGPT while preparing for the crime.
The Justice Department said the searches included questions about police response times, repairing the Llama Mini-Max .45 handg*n and obtaining ammunition despite his status as a convicted fel0n. His Google Maps history also showed a recent search for directions to the i3Bank.
The significance of the discovery was not that ChatGPT had somehow identified the robber. Rather, prosecutors were able to place the digital planning activity alongside physical and surveillance evidence already pointing toward Lewis-Payne.
The phone therefore became part of a much larger chain: the bank’s cameras captured his face; nearby cameras captured the vehicle; retail records tracked the spending; investigators recovered clothing and some of the cash; and the phone contained evidence of preparations for the robbery.
ChatGPT did not add 84 months to the sentence
The headline-grabbing role of AI can obscure what actually determined the punishment.
Lewis-Payne was not sentenced to 121 months because he used ChatGPT. The federal sentence arose from the underlying crimes: bank robbery and brandishing a firearm during a violent crime.
Federal law treats those offenses separately. Bank robbery by force or intimidation can carry a sentence of up to 20 years under 18 U.S.C. §2113(a).
Federal firearms law separately requires additional imprisonment when a firearm is used or carried during a crime of violence. Where the weapon is brandished, the statutory minimum is seven years, and that sentence is imposed in addition to the punishment for the underlying crime.
In Lewis-Payne’s case, that produced the 37-month robbery sentence followed by the 84-month firearms sentence.
The result was 121 months.
The case is also a reminder of how modern investigations work
The unusual part of the case is not simply that an accused bank robber consulted an AI chatbot.
It is that investigators were able to combine information from several digital systems that had little connection to one another at the time they were created.
A bank camera recorded a person. Residential cameras recorded a vehicle. Retail systems recorded purchases. Google Maps recorded a destination search. A phone preserved conversations with an AI system.
Individually, each piece of information offered only part of the picture. Together, they formed a detailed chronology of the robbery and its aftermath.
The pattern is increasingly familiar in criminal investigations: physical evidence is being supplemented by data generated as people move through ordinary digital services.
In this case, the technology intended to assist the suspect did not replace conventional police work. Surveillance footage, transaction records and recovered property remained central to the investigation. The AI conversations became another piece of evidence within that larger case.
A federal case with a distinctly modern trail
The underlying legal framework is decades older than generative AI.
The federal bank robbery statute dates to the 1930s, while the firearm provisions governing crimes of violence have also existed long before today’s conversational AI systems.
What has changed is the amount and variety of information available during an investigation.
Lewis-Payne’s case began with a conventional armed robbery in Omaha. By the time investigators had finished assembling the evidence, the investigation had touched bank surveillance, residential cameras, retail transactions, vehicle identification, Google Maps and an AI chatbot.
That combination is what makes the case notable.
The robbery itself lasted only long enough for a man to enter a bank, threaten employees and leave with cash. The digital trail created before and after it gave investigators considerably more time to reconstruct what happened.
