🔗 Share this article

Law enforcement agencies across the UK effectively campaigned to use a facial recognition system known to be biased against females, young people, and individuals from minority ethnic backgrounds, after complaining that a less biased version produced fewer potential suspects.

How the System Works

UK forces utilize the police national database (PND) to carry out retrospective facial recognition searches. This procedure entails comparing a reference photograph of a suspect against a repository of over 19 million mugshots to find possible hits.

Admitted Bias

The UK interior ministry conceded last week that the system was flawed. This admission followed a study by the National Physical Laboratory (NPL) determined it incorrectly matched people of Black and Asian heritage and females at much greater frequency than white men. The Home Office said it “took steps on the findings”.

“This raises the question of whether facial recognition only becomes effective if users tolerate biases in race and sex. Operational ease is a weak argument for disregarding fundamental rights.”

Known Issue

Official papers show that this bias has been known about for more than a year. Furthermore, law enforcement argued to overturn an initial decision that was designed to mitigate the problem.

Police bosses were notified of the algorithmic discrimination in late 2024. The Home Office-commissioned laboratory study concluded the system was had a higher probability to suggest incorrect matches for images depicting females, individuals of Black ethnicity, and those aged 40 and under.

A Reversed Decision

In reaction, the National Police Chiefs’ Council (NPCC) ordered that the confidence threshold required for potential matches be raised to a point where the disparity was significantly reduced.

However, this decision was overturned the following month following complaints from police that the adjusted system was generating fewer “useful lines of inquiry”. NPCC documents indicate the stricter setting reduced the proportion of searches that yielded potential matches from over half to a just under 15%.

Severe Disparities

Although the Home Office and NPCC refused to say what threshold is now in operation, the recent independent review found the system could produce incorrect matches for Black women nearly a hundred times more frequently than for Caucasian women at certain settings.

The ministry commented on these findings: “Our evaluation identified that in a specific scenarios the algorithm is more likely to incorrectly include some demographic groups in its search results.”

Operational Effectiveness vs. Bias

Outlining the effect of the brief increase to the system's confidence threshold, the police records state: “The change greatly lessens the impact of bias across legally safeguarded attributes of ethnicity, generation and gender but had a significant negative impact on operational effectiveness”. The documents further note that forces complained that “a once effective tactic now delivered outcomes of questionable value”.

Broader Rollout Plans

Meanwhile, the government has opened a ten-week public review on its plans to widen the use of facial recognition technology. Policing minister Sarah Jones has described the technology as the “most significant advance since genetic fingerprinting”.

Criticism from Advisors and Monitors

The chair of a police oversight board, head of the advisory panel for the police race action plan, said: “There was scant discussion through race action plan meetings of the facial recognition rollout even with obvious cross-over with the strategy's goals.

“This disclosure show yet again that the pledges to combat discrimination the police has made through the equality initiative are failing to be integrated into wider practice. Independent assessments have cautioned that innovative tools are being implemented in a landscape where ethnic inequalities, inadequate oversight and poor data collection already persist.

“All deployment of facial recognition must meet strict national standards, be subject to external review, and prove it diminishes rather than exacerbates ethnic bias.”

Official Statement

A government representative stated: “The Home Office takes the conclusions of the study seriously and we have already taken action. A new algorithm has been independently tested and procured, which has no statistically significant bias. It will be tested early next year and will be undergo evaluation.

“Our priority is protecting the public. This gamechanging technology will support police to apprehend and prosecute offenders. There is human involvement in every step of the process and no further action would be taken without trained officers carefully reviewing the results.”