- Woman in Oklahoma jailed for 6 months over crimes she denied committing
- Facial recognition software wrongly identified her as a fraud suspect
- The woman faced 16 charges, including 12 felonies, based on misidentification
A woman in the United States spent six months behind bars for crimes she insisted she never committed.
Kimberlee Williams, a resident of Oklahoma, was arrested after authorities in Maryland identified her as a suspect using facial recognition software. Investigators arrested her despite her repeated claims that she had never even set foot in the state, reported The Washington Post.
"I'm like, ‘How do I convince a detective that that wasn't me?" Williams said. "It's very obvious it's you," said an officer in Montgomery County, Maryland.
During a police interview in July 2021, she even offered to take a polygraph test and said her family could confirm she was not in Maryland at the time of the crimes. When she tried to explain, an officer cut her off, saying she wasn't telling the truth because of her past financial struggles.
“I'm not trying to waste your time, I'm telling you…” Williams said. "You're telling me, but you're not telling me the truth,” the officer interrupted.
As a single mother of seven, she began writing fraudulent checks decades ago to afford essentials such as diapers, which later extended to larger amounts of gifts. She pleaded guilty to misdemeanours in 2001 and 2004, was convicted of felony check fraud and conspiracy in 2010, and faced another fraud case in 2017.
However, in 2021, she made the decision to change her life, and by the end of that year, she was employed at a clinic.
While delivering a DoorDash order at Fort Sill in 2021, Williams was suddenly arrested over felony warrants from Maryland. The case traced back to December 2019, when a woman withdrew around $17,000 using fake checks from a SunTrust bank in Maryland.
A tip, reportedly based on facial recognition software, identified Williams as the suspect. Police in three counties, Montgomery, Prince George's, and Anne Arundel, charged her based on this identification and by visually comparing photos.
Williams ended up facing 16 charges, including 12 felonies, largely based on what appears to be a mistaken identity.
Later, it was revealed that a bank investigator used facial recognition technology to match her face with a suspect seen in bank fraud cases in Maryland. Based on this, the investigator told Montgomery County police that Williams was the person in the CCTV images.
When police filed charges against Williams, they did not tell the court that the identification came from facial recognition software.
Mitha Nandagopalan, an attorney for the Innocence Project, said, “When an officer seeks charges or seeks a warrant without revealing what they're basing their identification on, without revealing that it was facial recognition technology that produced this lead, they're essentially hiding or burying the unreliable step in that investigation.”
The American Civil Liberties Union (ACLU) stepped in and demanded an apology for Williams and introduced policy changes in how such facial technology is used. They also accused Montgomery police of malicious prosecution, saying that police knew facial recognition was used but still hid that fact while charging her.
