Nearly six months behind bars for a crime she could not have committed
Martin Mawyer
A Tennessee grandmother was arrested at gunpoint, jailed for months, and transported more than 1,200 miles from home, all because a facial recognition system got it wrong.
And no one stopped to question it.
In the summer of 2025, 50-year-old Angela Lipps of Elizabethton, Tennessee, was at home babysitting her grandchildren when U.S. Marshals arrived at her door with guns drawn. They informed her she was wanted for a series of bank fraud crimes in Fargo, North Dakota.
There was only one problem.
Angela Lipps had never been to North Dakota. Not once. Not ever.
A Computer Match Became the Case
Fargo police had been investigating multiple bank fraud incidents earlier that year. Surveillance footage showed a woman using a fake military ID to withdraw tens of thousands of dollars.
Investigators ran the footage through facial recognition software.
The system returned a match: Angela Lipps.
That match should have been treated as a lead, a starting point for further investigation.
Instead, it was treated as definitive.
According to reports, an investigator compared Lipps’ driver’s license and social media images to the suspect and concluded they were the same person based on facial features, body type, and hairstyle. That conclusion was used to obtain an arrest warrant.
From there, the system moved forward.
No One Verified the Most Basic Fact
At any point, authorities could have asked a simple question:
Was she even in North Dakota when the crimes occurred?
They did not.
Bank records later showed that while the fraud was taking place in Fargo, Lipps was in Tennessee, buying cigarettes and depositing Social Security checks, more than 1,200 miles away.
The evidence that would have cleared her existed all along.
It was never checked.
108 Days, Then More
After her arrest, Lipps was held in a Tennessee jail for 108 days without bail.
She was then extradited to North Dakota, where she remained in custody for nearly two additional months.
In total, she spent close to six months behind bars for a crime she could not have committed.
Only after her attorney obtained her bank records did the case fall apart.
The charges were dismissed.
But the damage had already been done.
Released With Nothing
On Christmas Eve, Angela Lipps was released from jail in Fargo.
She had no money. No coat. No way home.
Local defense attorneys helped her secure a hotel room and food. A nonprofit organization eventually assisted her return to Tennessee.
By that time, she had lost her home, her car, and even her dog.
According to reports, the Fargo Police Department has not issued an apology.
The Larger Problem
It would be easy to describe this as a technological error.
It is not that simple.
The facial recognition system made a suggestion. Everything that followed, the identification, the warrant, the arrest, the detention, was carried out by people.
At no point did the system pause long enough to verify the most basic facts.
At no point did someone stop to ask: What if we’re wrong?
Because if that question had been asked, it would have forced a second one.
What would it cost her if we are?
A Growing Pattern
This is not an isolated case.
In 2020, a Detroit man, Robert Williams, was wrongfully arrested after facial recognition software misidentified him in a theft investigation. The city later paid him $300,000 and revised its policies.
Cases like these are raising serious concerns about how facial recognition technology is being used in law enforcement.
Tools designed to generate investigative leads are increasingly being treated as if they provide definitive answers.
As artificial intelligence becomes more integrated into policing and other institutions, one question stands above the rest:
What safeguards exist between an algorithm’s suggestion and a life-altering decision?
In this case, those safeguards failed.
An algorithm produced a match.
And for nearly six months, no one questioned it.
Is This Case Really Over?
Despite the dismissal of charges, reports indicate that authorities have referred to Lipps as a “person of interest,” leaving open the possibility that charges could be refiled.
That means even now, after losing nearly everything, she remains tied to a crime she could not have committed.
For Interviews
Martin Mawyer, President of Christian Action Network, is available for interviews to discuss:
- The risks of facial recognition in law enforcement
- The growing pattern of AI-related wrongful arrests
- The lack of safeguards between AI outputs and real-world consequences
- What reforms are needed to protect civil liberties
Send media requests to: [email protected]
Stay Alert. Stay Informed.
Stories like this are not isolated. They are part of a broader shift in how technology is shaping law enforcement, freedom, and everyday life. For ongoing analysis and perspective, visit Christian Action Network’s Patriot Majority Report.