New reports reveal how technological biases have upended the lives of innocent people, raising urgent questions about the future of digital justice.
AI-driven facial recognition technology is considered one of the most significant technological advancements in the field of security. Although it has helped identify certain criminals, there are troubling aspects, such as some police departments relying excessively on the results of this technology without verifying sufficient evidence, sometimes leading to false arrests that harm the lives of innocent people. In this article, we highlight specific cases in the United States that demonstrate a notable absence of simple investigative steps and a reliance solely on what algorithms provide.
How did the story begin?
A suburb of St. Louis witnessed an assault on a security guard at a train station, resulting in serious injuries that affected his memory. After police failed to identify the suspects, investigators turned to a facial recognition program relying on hundreds of thousands of mugshots. Several suspect names appeared, including a person named Christopher Gatlin, who had no prior evidence linking him to the crime scene or records of violence. Although city policy states that facial recognition technology is “unscientific” and its results should not be the sole basis for an arrest, the case was built against Gatlin based on an unconfirmed technical match. He was held for more than two years before his innocence became clear.
Read also: The biggest artificial intelligence failures in 2024
Journalistic investigation findings reveal the scale of the case
A prominent American newspaper showed through the analysis of police documents, court reports, and interviews that at least 15 police departments across 12 states rely primarily on AI in some cases to identify suspects without supporting evidence. More dangerously, some officers described a facial match as a “100% match,” implying they treated the technology as a reliable reference without any additional investigation. Numerous cases have been documented of people wrongly arrested simply because facial recognition technology pointed to them, while crucial evidence such as eyewitness testimony, DNA data, or fingerprints was ignored.
Flaws in police implementation of the technology
Aspects leading to these wrongful arrests include neglecting to verify additional evidence, overreliance on witnesses, and technological bias against people with dark skin. Most police departments lack a policy obligating them to publicly report the use of facial recognition technology, and few keep adequate records of its use. This situation makes it difficult for lawyers and human rights advocates to track cases of arrest based on mistaken identification.
Read also: OpenAI research reveals high error rates in advanced AI models
The future and legal trends
With growing public and judicial awareness of the technology’s drawbacks, calls are mounting to establish stricter standards requiring police to verify additional evidence before proceeding with arrests. At the same time, some lawmakers are advocating for legislation that requires relevant authorities to fully disclose their reliance on algorithms in investigative processes. These steps would chart a fairer and more transparent path for the use of artificial intelligence in law enforcement.
Many are betting on artificial intelligence to enhance police efficiency and solve complex crimes, but recent events have proven that over-reliance on technology without sufficient safeguards can pose a threat to individual freedoms. Integrating digital evidence with traditional investigation and rigorous verification of witnesses and physical evidence is the ideal way to ensure justice is served, so that innocent people do not pay the price for smart software errors.
