The key UK CCTV and surveillance statistics in one place — how many cameras there are, which council and police areas are the most watched, how fast police facial recognition is scaling, and what the law requires of camera operators, with the data period and source stated next to every figure.
The figures below come from the Home Office’s December 2025 factsheet, the Met’s September 2025 annual report and May 2026 pilot results, and Comparitech’s original FOI research updated on 28 May 2026. The FOI study is independent research, not official national statistics, and covers disclosed council and police cameras rather than all UK cameras.
This page brings the headline numbers from each of those sources together. It is scoped to the deployment and usage of physical surveillance — cameras, facial recognition, ANPR and body-worn devices — plus the UK GDPR duties that apply to camera operators. Public attitudes to automated decision-making are covered separately on our AI and data privacy statistics page.
Key facts and figures
- 3,147,436 estimated faces passed Met live facial recognition cameras across 203 deployments from 11 September 2024 to 10 September 2025; 962 people were arrested.
- 13 police forces had deployed or were using live facial recognition as of November 2025.
- 25,000+ retrospective facial-image searches are run against the Police National Database every month.
- 10 false alerts were recorded among 2,077 alerts in the Met’s 2024/25 reporting period: 0.48% of alerts and about 0.0003% of estimated faces seen.
- 13.8 cameras per 1,000 residents was North Ayrshire’s rate in the FOI study updated May 2026, the highest among responding councils (1,849 cameras).
- 130,000 cameras, approximately, were disclosed by 35 responding police forces in the study updated May 2026; the Met reported 31,077.
- 26,500 body-worn cameras were reported for the Met in that study — about 85% of its disclosed camera total.
- 173 arrests were reported during the Met’s 24-operation Croydon pilot, October 2025–March 2026.
Checked on 15 October 2026. Each figure retains the date of its report or observation period. Camera inventories, police deployments and research samples are different measures; none supplies a current census of every UK surveillance camera.
How many CCTV cameras are there in the UK?
There is no verified current national camera census in the sources used here. Published totals often mix fixed CCTV, body-worn devices, private cameras and older estimates. A camera count also does not measure how often a person is recorded, how long footage is retained or how intrusive the processing is.
Comparitech’s study, updated on 28 May 2026, sent requests to 380 councils and 45 police forces. Thirty-five forces provided figures, two for body-worn cameras only. They disclosed about 130,000 cameras, including 31,077 for the Met. The figures exclude undisclosed cameras and the private estate and should not be extrapolated to a national total.
The area comparisons below use the study’s disclosed counts and population denominators. Cameras recording identifiable people can process personal data, but the applicable rules depend on purpose and operator: ordinary workplace surveillance, domestic use and police law enforcement do not all fall under identical provisions.
Which responding areas had the highest camera densities?
In the FOI study updated May 2026, North Ayrshire had the highest council-camera rate among respondents: 1,849 cameras or 13.8 per 1,000 residents. Hackney’s 3,281 cameras equated to 13.2 per 1,000. These are disclosed council devices, not all public and private cameras in either area.
The study reported council-camera rates of 3.6 per 1,000 in Scotland, 1.9 in England, 2.4 in Wales and 2.5 in Northern Ireland. Its police comparison put City of London at 48.9 per 1,000 residents. Resident-based ratios can be misleading in areas with many commuters and visitors, and do not establish how intensively individuals are monitored.
| Area | Cameras per 1,000 people | Total cameras | Detail |
|---|---|---|---|
| North Ayrshire | 13.8 | 1,849 | Highest council rate among respondents |
| Hackney | 13.2 | 3,281 | Highest English council rate among respondents |
| City of London (police area) | 48.9 | 738 | Police cameras; small resident denominator |
| Scotland (council average) | 3.6 | — | Study’s disclosed council cameras |
| England (council average) | 1.9 | — | Study’s disclosed council cameras |
These are the figures in the study updated on 28 May 2026, covering disclosed cameras only. Counts and population denominators may have different reference dates.
Use the table to compare the reported inventories, with its coverage limits. A higher camera-to-resident ratio does not by itself demonstrate more effective crime prevention or greater privacy impact.
How many police forces use live facial recognition?
13 police forces had deployed or were using live facial recognition (LFR) as of November 2025, according to the Home Office factsheet updated on 4 December 2025. LFR works by comparing faces captured on a live camera feed against a watchlist in real time; alerts are only generated for a possible match, and the factsheet states that images of people not on the watchlist are deleted immediately and automatically.
Retrospective facial recognition compares a still image with a database after an event; live recognition compares passing faces with a watchlist; operator-initiated recognition supports an officer’s identity check. The December 2025 Home Office factsheet reported more than 25,000 monthly facial-image searches on the Police National Database. This is a dated search volume, not a count of people arrested or a directly comparable volume of live scans.
Police processing for law-enforcement purposes is generally governed by Part 3 of the Data Protection Act 2018, including its rules for sensitive processing, rather than the ordinary UK GDPR Article 9 regime. High-risk processing requires an impact assessment. Necessity, proportionality, equality and human-rights duties also matter; other police administrative processing can fall under UK GDPR.
How accurate is police facial recognition?
The Met’s annual report covers 11 September 2024 to 10 September 2025. It recorded 203 deployments, 2,077 alerts, 10 false alerts and an estimated 3,147,436 faces seen, with 962 people arrested. Ten divided by all alerts is about 0.48%; divided by estimated faces seen it is about 0.0003%. These denominators answer different questions and do not measure every possible missed match.
The December 2025 Home Office factsheet describes NPL tests at specified settings: the retrospective system found the correct match in 99% of searches where that match was in the database; live recognition identified 89% of watchlisted people, with a worst tested false-alert probability of about one in 6,000 for a 10,000-image watchlist. Test settings and image conditions matter, so none is a universal accuracy guarantee.
The Met reported 173 arrests from 24 Croydon pilot operations between October 2025 and March 2026, equivalent to one per 35 deployment minutes. More than 470,000 people passed the cameras and one false alert led to a brief interaction, not an arrest. The force also reported a 10.5% local crime reduction and a 21% reduction in offences classified as violence against women and girls versus the prior-year period. Those before-and-after comparisons alone do not isolate the effect of facial recognition.
What about ANPR and body-worn cameras?
The FOI study reported 26,500 Met body-worn cameras, around 85% of its disclosed total, and 6,176 for Police Scotland. Including portable devices explains why these inventories should not be described as counts of fixed street CCTV.
Twenty-four police forces supplied ANPR counts in the study. Sussex reported the highest among those forces, at 599, followed by the Met at 450; this was not the highest council-or-police count, because Bradford Council reported 650. Number plates, location and time can be personal data when linked to identifiable people. Police law-enforcement uses and ordinary commercial uses have different statutory frameworks.
Is workplace CCTV legal under UK GDPR?
Yes — workplace CCTV is legal, but only if it meets the UK GDPR's conditions. An employer running cameras is a data controller processing personal data, so it needs a lawful basis (usually legitimate interests), clear signage telling people they are being recorded, a Data Protection Impact Assessment where the monitoring is high-risk, a defined retention period, and controlled access to the footage. The ICO's CCTV and video surveillance guidance sets out those expectations in detail.
Covert staff monitoring requires exceptional justification and a proportionate, tightly controlled approach. Biometric recognition used to uniquely identify someone requires an appropriate special category condition as well as a lawful basis under UK GDPR, and is likely to require a DPIA. Ordinary identifiable CCTV footage is not automatically special category biometric data merely because it shows a face.
Frequently asked questions
How many CCTV cameras are there in the UK?
The sources here do not establish a current national total. The FOI study updated May 2026 reported around 130,000 cameras across 35 responding police forces, including 31,077 for the Met, but excluded undisclosed and private cameras. Its totals also include body-worn devices.
Which council had the highest camera density in the FOI study?
North Ayrshire had the highest rate among responding councils in the study updated May 2026: 13.8 per 1,000 residents, from 1,849 cameras. This is a disclosed council-camera ratio, not a measure of all surveillance in the area. Hackney’s equivalent rate was 13.2.
How many police forces use live facial recognition?
13 forces had deployed or were using live facial recognition as of November 2025, per the Home Office factsheet of 4 December 2025. Police also run more than 25,000 retrospective facial-image searches against the Police National Database each month.
How accurate is police live facial recognition?
In the Met’s 11 September 2024–10 September 2025 report, there were 10 false alerts among 2,077 alerts and an estimated 3.15 million faces seen. That is about 0.48% of alerts or 0.0003% of faces seen. Operational and laboratory measures use different settings and denominators and do not establish a universal accuracy rate.
Is workplace CCTV legal under UK GDPR?
Workplace CCTV must be necessary and proportionate, have a lawful basis, provide suitable information, limit retention and access, and have a DPIA where likely to create high risk. Biometric identification and covert monitoring need additional assessment. A camera image alone is not automatically special category data.
Do ANPR cameras count as surveillance under GDPR?
ANPR information can be personal data where it relates to an identifiable person. Commercial operators normally assess UK GDPR duties; police law-enforcement processing generally falls under Part 3 of the Data Protection Act 2018. The FOI study’s 599 Sussex cameras was the highest among 24 responding police forces, not all UK camera operators.
Related guides
- GDPR and CCTV in the UK: rules for businesses and homeowners
- AI and Data Privacy Statistics UK: attitudes and adoption
- ICO Complaints Statistics UK: what people complain about
- Data Breach Statistics UK: ICO reports, trends & costs
- What is personal data under GDPR?
Sources & references
- Home Office — police facial recognition factsheet, December 2025
- Comparitech — original council and police FOI study, updated May 2026
- Metropolitan Police — LFR Annual Report, September 2025
- Metropolitan Police — Croydon pilot results, May 2026
- ICO — CCTV and video surveillance guidance
- ICO — law enforcement processing under Part 3 of the Data Protection Act 2018
Every camera that records identifiable people brings UK GDPR duties — train your team to handle CCTV, ANPR and monitoring data lawfully.
Explore the GDPR & Data Protection Course →
Mark writes about data protection, workplace compliance and accredited online training for GDPR & Data Protection Course, part of Online CPD Academy.