UK Immigration Facial Recognition Trials Logged 10,264 Face Sightings

The UK's published Immigration Enforcement facial-recognition log records 10,264 face sightings across 18 deployment entries from November 2025 through February 2026, a TFNOS calculation from the Home Office's transparency table shows. The same entries record three outcomes in a column labelled "Arrests or disposals". Neither figure establishes the number of unique people scanned or proves that facial-recognition matches caused the recorded enforcement outcomes.

The government updated its deployment-results page on 15 October 2026. Adding the values in the table's "Faces seen" column gives 7,512 across eight entries dated November 2025 and 2,752 across ten entries dated February 2026. The first November entry covers three days and 10 ferry sailings, so a row is not necessarily one short camera operation. These totals are face sightings recorded by the system, not distinct individuals.

The table reports one arrest or disposal in a November 28, 2025 entry and one each in two February 25, 2026 entries. In both February entries, the "Total alerts" column is zero even though "Positive interventions" and "Arrests or disposals" each record one. The public table therefore does not justify describing the three outcomes as arrests caused by facial-recognition matches. The Home Office should clarify how non-alert interventions and disposals are counted.

What the system is intended to do

The Home Office says live facial recognition compares camera images with a predetermined watchlist in real time and generates alerts for possible matches. Its published policy and legal documents describe restrictions on watchlists, authorisation, human review and data retention. The deployment log itself does not reveal the identities on watchlists, the basis for individual inclusion or whether later operational safeguards were followed.

The numerical record matters as the department considers broader uses of the technology. Computer Weekly reported on 30 October that updated Immigration Enforcement documents contemplate use in civil immigration matters and locations beyond ports of entry. Its report included a Home Office defence of the programme as a tool for immigration enforcement and public protection, as well as concerns about the reliability of underlying identity records. Prospective policy provisions are not evidence that a new deployment has occurred at a particular inland location.

The Home Office says facial recognition does not create new enforcement powers and that officers, not the technology, make decisions on stops, detention or arrest. Critics question the proportionality of scanning people not suspected of wrongdoing and the quality of information used to construct watchlists. The published figures do not settle those issues.

What remains unknown

Useful further disclosures would include a deployment-by-deployment explanation of the three arrest-or-disposal entries, whether an alert contributed to each outcome, and current authorisation and watchlist audit records. Later deployment results would also show whether the frequency or effectiveness of operations changed after the October policy revisions.

TFNOS has not established that anyone was wrongly identified, unlawfully detained or included on a watchlist without legal authority. This comparison is confined to the official numerical record, published policy and questions the documents leave open.