Tennessee grandma jailed for 5 months after AI flagged her for bank fraud in state she never visited

A Tennessee grandmother found herself in a nightmare situation when a facial recognition program wrongly identified her as a suspect in a bank fraud case in a state she had never even visited. Angela Lipps, 50, was arrested at her rental home in Tennessee in July and was later extradited to Fargo, North Dakota, over 1,000 miles away from her residence.

The West Fargo Police Department utilized facial recognition technology that flagged Lipps as a potential suspect in a local fraud case. Despite the error, the department proceeded with further investigations to confirm Lipps’ identity as a suspect. It was later revealed that the department used Clearview AI, which mistakenly identified Lipps due to similarities in facial features.

Lipps, who had never been on an airplane before, was detained in Tennessee for three months due to a miscommunication between authorities. Her extradition waiver was not relayed to North Dakota, leading to a prolonged period of uncertainty and distress for Lipps.

After finally arriving in Fargo, Lipps was able to prove her innocence with bank records showing that she was in Tennessee at the time of the alleged fraud. Charges against her were eventually dismissed on Christmas Eve, marking the end of a five-month ordeal.

Throughout her detention, Lipps suffered significant personal and financial losses. Her reputation was tarnished, her rental home was lost, and her belongings were seized when her storage unit bill went unpaid. The emotional toll of the experience left Lipps feeling like a changed person.

A GoFundMe campaign was set up to support Lipps, raising $68,000 to help her rebuild her life after the wrongful arrest. In response to the incident, the Fargo Police Department announced that they would no longer use information from Clearview AI and would implement stricter oversight of facial recognition technology.

The department acknowledged their mistakes in handling Lipps’ case and pledged to improve their protocols for identifying suspects in fraud cases. They are currently working to rectify the situation and prevent similar errors in the future.