Police Department Testing AI-Powered Detective on Real Crimes

Hello!

A police department in the United Kingdom is testing an AI-powered system that could help solve cold cases by condensing decades of detective work into mere hours, Sky News reports.

Developed in Australia, the platform known as Soze is currently being evaluated by the Avon and Somerset Police Department. The system scans and analyses large volumes of evidence, including emails, social media accounts, video footage, financial statements and other documents. While the tool shows promise, its accuracy rate has not yet been disclosed—an important consideration given that AI models can produce incorrect results or hallucinate information.

Accelerating Investigations

Sky reports that Soze processed evidence from 27 complex cases in roughly 30 hours, a workload estimated to require 81 years of human effort. For police forces facing personnel shortages and budget pressures, such efficiency gains are particularly attractive.

“You might have a cold case review that just looks impossible because of the amount of material there and feed it into a system like this which can just ingest it, then give you an assessment of it,” Gavin Stephens, chair of the UK’s National Police Chiefs’ Council, told Sky. “I can see that being really, really helpful.”

Minority Report

Stephens also highlighted another AI initiative: the creation of a national database of knives and swords used in attacks across the United Kingdom.

Although Stephens is optimistic about the near-term deployment of these tools, experts stress the need for thorough validation. AI systems used in law enforcement have repeatedly shown vulnerability to errors and bias. One predictive model intended to assess the likelihood of reoffending was found to be both inaccurate and biased against Black individuals—echoing concerns raised by Philip K. Dick’s novella “Minority Report” and its 2002 film adaptation.

AI facial recognition technology has likewise led to wrongful arrests, disproportionately affecting minority communities. These risks prompted the US Commission on Civil Rights to criticise the use of AI in policing.

Despite the perception that machine-driven analysis is inherently objective, AI systems are trained on data collected and labelled by humans. As a result, existing biases and errors can be embedded in the technology from the outset.

Also read:

• The Power of POS: Streamlining Business Operations for Success
• People Are Apparently Microdosing Ozempic
• How To Cancel Audible Membership

Thank you!
Join us on social media!
See you!