On November 4, 2026, the UK Home Office, announced by Home Secretary Shabana Mahmood, set out plans to replace what it called the “vast patchwork system” governing police facial recognition with a clearer set of laws, following a public consultation. A specific law will require police use to be “intelligence led and proportionate”, deployed only where officers believe wanted people could be present.
Any technology used by law enforcement must be “rigorously and independently tested” for bias based on age, gender or race. A new independent regulator will oversee current and future police technologies, advise on standards, scrutinise use and make recommendations to ministers, and in extreme cases can seek a court injunction to stop a force using a technology. A legal code of practice for police will be drawn up with the regulator, and a separate code will cover private-sector use, including shops using facial recognition against crime.
The government said the consultation showed strong support for the technology but a need for “stronger safeguards around the extent of its use”. It cited 173 arrests from a Croydon pilot between October 2025 and March 2026 and announced 11 million pounds to triple the number of live facial recognition vans available to police.
Why it matters: live facial recognition is one of the most visible uses of AI by the state. The UK is choosing to expand it while putting it on a statutory footing with a regulator that can halt deployments, a different path from outright bans adopted by some US cities.
What it does not show: no bill has been published and the announcement gives no date for legislation or the codes. The arrest and crime figures come from the government and the Metropolitan Police, not from independent evaluation, and the bias-testing standard has not been defined.