UK Immigration Face-Scanning Log Shows 10,264 Sightings and Three Outcomes

UK Immigration Enforcement's published facial-recognition deployment records list 10,264 face sightings across 18 entries in November 2025 and February 2026, a TFNOS calculation from the official table shows. The same table records three outcomes under the heading "Arrests or disposals". The figures do not establish how many distinct individuals were seen or whether a facial-recognition alert led to any of the three outcomes.

The Home Office updated its transparency record on 15 October 2026. Adding the "Faces seen" values yields 7,512 across eight entries from November 2025 and 2,752 across ten entries from February 2026. The first November entry spans three days and ten ferry sailings, so the 18 table entries should not be described as 18 separate single-session operations. A person could have appeared more than once.

One November 28 entry and two February 25 entries each record one arrest or disposal. Both February entries list zero total alerts but one positive intervention and one arrest or disposal. That discrepancy in column definitions is a reason to ask for clarification, not evidence of an unlawful operation. In particular, the public table does not prove that three people were arrested because a face match identified them.

Wider policy context

The Home Office says live facial recognition compares live camera feeds against approved watchlists, generating alerts for trained officers to assess. It says decisions on detention, arrest and removal must be taken by officers exercising existing legal powers. The department's public documentation addresses watchlist controls, authorisation, human oversight and deletion.

Computer Weekly reported on 30 October that updated Immigration Enforcement policy documents contemplate use in civil immigration matters and locations beyond ports. The report carried the Home Office's position that the programme helps enforce immigration rules and protect the public, while recording concerns about data quality and whether people not accused of criminal offences could be scanned. The documents do not establish that any particular new inland deployment has occurred.

The 10,264 total counts face sightings, not unique people or confirmed matches. Likewise, the three arrests-or-disposals values are a category in an administrative table, not a verified causal effectiveness rate for the biometric software. A defensible comparison would require the deployment-level alert records and explanations of the outcomes.

Records still needed

The Home Office should clarify the two February entries with positive interventions but no alerts, identify whether the three outcomes arose from facial matches or separate officer checks, and publish updated deployment statistics, authorisation logs and watchlist governance information. These would allow meaningful scrutiny of proportionality, effectiveness and error controls.

This analysis does not allege that any individual was wrongly identified, unlawfully detained or placed on a watchlist without authority. It reports what the public records show and distinguishes those facts from unresolved questions.