A 50-year-old grandmother from Tennessee was in jail for more than five months after police used an AI facial recognition tool to charge her with crimes she denies committing.
Angela Lipps was arrested on July 14 after a warrant from North Dakota appeared in the police system. The charges were related to a string of bank fraud cases in and around Fargo, more than 1,000 miles from Lipps’ home.
The police used facial recognition technology in their investigation to find the suspect. A police department in the area used software from Clearview AI, which compares an image to a large database of other images collected from the internet. The software identified Lipps as a possible match.
That lead moved forward. Police combined it with what they described as other investigative steps. Still, it remains unclear what other evidence tied Lipps to the crimes.
How AI and Oversight Gaps Led to a Wrongful Arrest
Fargo Police Chief Dave Zibolski later admitted that the department made “a few errors” in the process. He said officers relied on information from another agency’s AI system without full oversight. He also said leadership did not know that the partner agency had adopted that system.
The West Fargo Police Department had used the AI tool to identify a “potential suspect” based on a fake ID image. They shared that result with Fargo police. However, they did not file charges themselves and said they did not have enough evidence to do so.
Lipps’ arrest set off a long chain of events. She spent more than three months in a Tennessee jail before extradition. The delay remains unclear. Officials have not confirmed whether it was due to legal disputes, paperwork gaps, or other issues.
In October, she was sent to North Dakota. She faced multiple felony charges, including theft and misuse of personal information.
Her experience during extradition was distressing. She said it was her first time on a plane. She felt scared and humiliated.
Once in Fargo, her legal team began to review the case. They found bank records that showed she was in Tennessee at the time of the fraud incidents. This evidence pointed away from her involvement.
The Case of Porcha Woodruff Lipps
By mid-December, prosecutors acknowledged that the defense had provided evidence that could clear her. On December 23, the court dismissed the charges without prejudice. This means the case can reopen if new evidence appears. Lipps was released on Christmas Eve.
Her lawyers say the damage runs deep. She lost her freedom for months. Her reputation suffered. The emotional toll remains.
They also question why basic checks did not happen earlier. Police knew she lived in Tennessee. Her attorneys argue that a simple review of travel or financial records could have raised doubts before her arrest.
Fargo police now say they will change their procedures. They will no longer use data from the partner agency’s AI system. Instead, they plan to work with trained state and federal units that specialize in facial recognition.
They will also review all facial recognition cases each month. This step aims to add oversight and reduce errors.
Another key issue involves communication gaps. Police said there is no simple system to alert them when a suspect is in custody in another state. They are now considering daily checks of jail records to fix that problem.
Despite these steps, the department has not issued a direct apology. Officials say the case remains open, and they are still reviewing who may be involved in the fraud network.
The Lipps Case and the Perils of Facial Recognition
Experts say this case reflects a wider problem. Police departments across the country are adopting AI tools at a fast pace. Many rely on vendor claims rather than strong evidence of accuracy.
Facial recognition can help solve crimes, but it carries risk. Errors can occur when systems match the wrong face. These errors can grow when officers trust the technology without careful review.
Researchers note that most failures come from a mix of human and technical mistakes. Officers may treat AI results as firm leads rather than rough suggestions. That shift can lead to weak investigations.
The Lipps case shows how that risk plays out in real life. A single match from an AI system helped trigger an arrest, detention, and cross-state transfer. Later, simple records helped undo the case.
Her lawyers are now exploring civil rights claims. They have not filed a lawsuit yet.
For Lipps, the legal outcome brings some relief. Still, the experience has changed her. She says she has no plans to return to North Dakota.
The case leaves a clear question: how should police balance new tools with basic investigative work? For now, it stands as a warning about the cost of getting that balance wrong.
