Grandmother Jailed 6 Months After AI Misidentified Her

Angela Lipps, a 50-year-old grandmother from Elizabethton, Tennessee, spent nearly six months in jail because a facial recognition algorithm said she robbed a bank in Fargo, North Dakota - a city she has never visited, 1,200 miles from her home.

Bank records proved she was buying cigarettes and depositing Social Security checks in Tennessee at the exact times police claimed she was committing fraud in Fargo. Nobody checked those records for five months.

TL;DR

Angela Lipps, 50, mother of three and grandmother of five, arrested at gunpoint on July 14, 2025 while babysitting four children
Fargo PD used facial recognition to match her to surveillance footage from a bank fraud case
She sat in a Tennessee jail for 108 days before North Dakota officers even picked her up
First police interview: December 2025, five months after arrest
Charges dismissed Christmas Eve after bank records proved she was in Tennessee the entire time
She lost her home, her car, and her dog. No apology from Fargo PD

The Crime She Didn't Commit

In April and May 2025, a woman walked into multiple Fargo banks with a fraudulent US Army military identification card and withdrew tens of thousands of dollars. Surveillance cameras captured the transactions. Fargo Police Department ran the footage through facial recognition software. The algorithm returned a match: Angela Lipps of Elizabethton, Tennessee.

That was the entire basis for the arrest. No fingerprints. No witness identification. No phone location data. No travel records. One algorithmic match.

The Arrest

On July 14, 2025, US marshals arrived at Lipps' home. She was babysitting four children. They arrested her at gunpoint.

Lipps told WDAY News that no one from the Fargo Police Department ever called to question her before the arrest. No detective contacted her to verify her identity, check her whereabouts, or confirm that a 50-year-old grandmother from rural Tennessee had any connection to bank fraud in North Dakota.

She was booked into a Tennessee county jail as a fugitive and held without bail.

108 Days in a Tennessee Jail

Lipps sat in the Tennessee jail for 108 days. During this time:

No Fargo detective interviewed her
No one checked her bank records
No one verified her location during the alleged crimes
No one compared her physical appearance to the actual suspect beyond the algorithm's match

North Dakota officers didn't pick her up until October 30, 2025 - more than three and a half months after her arrest. It was the first time anyone from the investigating department had face-to-face contact with her.

The Evidence That Cleared Her

When her North Dakota attorney, Jay Greenwood, finally obtained her bank records, the case collapsed instantly. The records showed that during the exact dates and times of the Fargo bank frauds, Lipps was in Tennessee:

Depositing Social Security checks at her local bank
Buying cigarettes at a gas station
Buying pizza
Using Cash App to order Uber Eats

She was 1,200 miles away. The proof had been available from the first day. No one looked.

"If the only thing you have is facial recognition, I might want to dig a little deeper."
Jay Greenwood, attorney

Christmas Eve

On December 19, 2025, prosecutors met with Greenwood and reviewed the bank records. On Christmas Eve, the charges were dismissed and Lipps was released.

By then she had been incarcerated for around 164 days. She had lost:

Her home - unable to pay rent while jailed
Her car - repossessed
Her dog - given away while she was incarcerated

Local defense attorneys in Fargo funded a hotel room for her after release. The F5 Project, a North Dakota nonprofit, arranged her transportation home to Tennessee. A West Fargo resident created a GoFundMe to help her rebuild.

The Fargo Police Department hasn't apologized. It didn't offer to cover her travel expenses home.

Where It Falls Short

This Is Not an Isolated Case

Angela Lipps is at least the seventh documented case of wrongful arrest based on facial recognition in the United States:

Trevis Williams (NYPD, April 2025) - Arrested despite being over 6 inches taller than the suspect in surveillance footage
Porcha Woodruff (Detroit, 2023) - Arrested while 8 months pregnant based on a facial recognition match to a carjacking suspect
Robert Williams (Detroit, 2020) - Arrested in front of his children on his lawn for a shoplifting case
Nijeer Parks (New Jersey, 2019) - Spent 10 days in jail and nearly took a plea deal for a crime committed 30 miles from his home

Every documented wrongful arrest from facial recognition in the US has involved a Black person - until Lipps, who's white. The technology's error rate is highest for women and people of color, but the Lipps case shows that the institutional failures surrounding it - no verification, no interview, no corroboration - affect everyone.

The Real Problem Is Not the Algorithm

Facial recognition software is explicitly designed to produce candidate matches, not positive identifications. Every vendor's documentation states that results must be verified by human investigation before any law enforcement action. The algorithm is a lead, not evidence.

In Lipps' case, Fargo PD treated the algorithmic output as sufficient for an arrest warrant. They didn't:

Interview the suspect
Check travel records
Verify financial records
Compare physical characteristics beyond the facial match
Contact Tennessee law enforcement for assistance

The failure is not that the AI made a mistake. AI facial recognition produces false matches routinely. The failure is that a police department used an unverified algorithmic output to jail a woman for six months without performing basic investigative steps that would have cleared her in hours.

A facial recognition algorithm said Angela Lipps robbed a bank in a city she has never been to. Police arrested her at gunpoint while she was babysitting. She spent 164 days in jail. Bank records that would have cleared her in an afternoon weren't checked for five months. She lost her home, her car, and her dog. The charges were dismissed on Christmas Eve. Nobody has apologized. The algorithm is still in use.

Sources: