MPS confirms the use of AI to spot potential misconduct
The Metropolitan Police Service (MPS) has confirmed it is using AI tools to analyse internal sickness and absence data in a bid to identify failings in “standards, culture and behaviour”.
The software, built by US data analytics firm Palantir Technologies, is designed to identify patterns in behaviour that might seem innocuous in isolation but, when aggregated, could potentially point towards misconduct.
The company’s name comes from the seeing-stones in JRR Tolkien’s The Lord of the Rings, magical orbs that allow the dark lord Sauron to observe events across vast distances.
Data points used by the software include instances of accessing files unrelated to assigned cases, changes in financial circumstances or the filing of large numbers of public complaints, regardless of whether they are upheld.
Whereas vetting previously consisted of a snapshot in time, the new system introduces a form of continuous vetting, creating a risk score for each officer. Those exceeding a pre-determined threshold are then flagged to the Directorate of Professional Standards.
Speaking to The Guardian, The Police Federation criticised the approach as “automated suspicion”. It said: “Officers must not be subjected to opaque or untested tools that risk misinterpreting unsustainable workload pressures, sickness or overtime as indicators of wrongdoing.”
“Any system that profiles officers using algorithmic patterns must be treated with extreme caution. Policing already operates under some of the broadest and deepest scrutiny of any profession … If forces are serious about raising standards and public confidence, the focus must remain on proper supervision, fair processes and human judgment, not the automation of suspicion.”
A spokesperson for the MPS told Police Professional: “We are working with Palantir on a time-limited pilot as part of our professional standards work.
“The pilot brings together internal data we already hold from multiple standalone systems into a form which can be used without delay. This is because there is evidence to suggest a correlation between significant levels of sickness, increased absences or unusually high overtime, and failings in standards, culture and behaviour.
“The aim of the pilot and tool is to help us identify these patterns of behaviour in our staff. This is part of our wider effort to drive up standards and improve the Met’s culture.”