SIA Licence Holders Fall by More Than 10,000 After Four Years of Growth

Active SIA licence holders fell from 457,694 in September 2025 to 447,558 in September 2026. That is a net reduction of 10,136 people, a decline of 2.2%. It follows four consecutive annual increases across the September snapshots from 2021 to 2025.

These figures count people who held a valid SIA licence at a specific point in time — not people in active employment. A fall in licence holders does not establish how many people left security work.

This article examines that reversal against the five-year trend, looks at where licence numbers fell by category, and considers what the application data adds — and what it cannot establish.

The five-year picture: a decline after substantial expansion

To understand whether the 2026 figure represents a meaningful shift, it helps to look at the run of September snapshots since 2021.

Two things stand out. First, the licensed population grew for four consecutive years, at varying rates, before this year’s fall. Second, despite the 2026 decline, the total remains 62,646 higher than in 2021 — an increase of 16.3% over five years.

September 2025 was the highest count in this series, the peak of a sustained expansion. The 2026 figure is below that September 2025 peak but remains above the 2021 level.

Where did active licence numbers fall?

Moving from people to licence categories requires a change in what is being measured — one person can hold more than one licence, so category totals do not map directly onto headcount.

The strongest finding here is the split between the two largest categories. Door-supervision licences fell by 2.9%, the largest numerical decline by a wide margin, while CCTV licences grew by 5.5%. Security-guarding licences held broadly flat.

This is a meaningful pattern worth investigating further, but it should not be read as evidence that door supervision alone explains the fall in individual licence holders.

These tables cannot show which specific individuals gained, lost, or retained different licence types. A person could have let a door-supervision licence lapse while keeping a security-guarding licence active.

The category data is indicative of where activity is concentrated, not a full explanation of the headcount change.

First-time applications also fell — but they measure something different

Application data for January to August adds a further layer.

First-time paid applications — those from applicants who have never previously held an SIA licence — fell by 11,484, a drop of 20.9%. Repeat applications — from those who have held a licence before — rose by 6.1%.

That divergence is the sharpest signal in the application data: First-time paid applications fell, while repeat paid applications from people who had previously held an SIA licence increased. Three qualifications bear on how far this can be taken:

• The licence-holder change covers September 2025 to September 2026. The applications comparison covers January to August 2025 against January to August 2026. These are related but different periods, and the application figures should not be read as a direct accounting explanation for the September-to-September decline. A closer comparison would use paid applications over September to August in consecutive years.
• Paid applications cannot be treated as a count of licences granted. The two measure different things.
• Repeat applications are not exclusively renewals and cannot establish a renewal rate — that would require the number of licence holders eligible or due to renew as a denominator, which the applications workbook does not publish.

What could explain the reduction?

Several documented changes provide relevant context. The sources below record real pressures on applicants and licence holders, but none establishes how much, if any, of the 10,136 reduction it caused.

Training and the cost of renewal

Refresher training became compulsory from April 2025. The SIA introduced a refresher qualification requirement for door-supervisor and security-guard renewals from that date.

A first-aid qualification had already been required before April 2025. Together, these requirements add time, planning, and cost to the renewal process, and are worth examining as factors in renewal decisions, particularly for occasional workers.

However, security-guarding licences rose slightly over the same period that door-supervision licences fell — two categories that moved in different directions despite both being subject to the refresher requirement.

These totals alone cannot isolate the effect of training. Different renewal volumes, entry rates, and other factors could produce different overall results across categories.

The £20 licence-fee rebate ended in April 2026. The SIA confirmed on 4 March 2026 that its temporary rebate would not continue, raising the amount paid per application from £184 to £204.

The statutory fee itself did not rise — the discount ended because the historical reserves funding it were exhausted. This increased the upfront cost during part of the period covered by the latest figures.

It cannot, however, explain the full January–August application decline: monthly data already shows lower first-time applications in January, February, and March 2026, before the rebate ended.

How applicants fund their licences

The SIA’s strategic plan 2026–2029 reports survey findings from 2025. Among 12,000 first-time applicant respondents, 75% paid for their own licence.

Among 8,434 renewal respondents, 63% paid personally, with a further 12% repaying an employer advance through wages. These are survey findings rather than universal figures.

They strengthen the case for treating affordability as a genuine factor worth examining, particularly when combined with the end of the rebate and the additional cost of mandatory refresher training.

Working conditions and retention

Get Licensed’s UK Workplace Violence Report 2026 surveyed 537 security professionals in January and February 2026. It reported that 33% had taken time off, or considered doing so, because of violence they personally experienced.

The sample was drawn from people who had recently trained with Get Licensed, so it should not be read as representative of the wider licensed population.

It offers evidence for the strain of frontline work, but the survey does not connect that strain to any specific change in the active licence-holder population.

For a practitioner’s perspective on these pressures, Tom Richmond’s commentary examines workplace violence and investment in retail security staff.

What the SIA says about renewals and processing

The SIA’s strategic plan 2026–2029 states that renewal rates had remained around 50%, with applications processed in fewer than ten working days on average, including complex cases. Both figures are useful context.

A reported average of fewer than ten working days does not rule out longer waits for particular applicant groups or case types, and the renewal rate of around 50% does not reveal category-specific rates or the number of licence holders due to renew during the comparison period.

Neither figure settles what happened in the months immediately before the September 2026 snapshot, and both predate it.

That leaves three important lines of investigation running alongside each other: the fall in first-time applications, the volume of returning holders entering the active population, and the volume of people losing their last active licence.

None can be confirmed as the leading cause without further analysis.

Does this mean employers face a staffing shortage?

Not necessarily, and not automatically. Licence-holder numbers measure potential licensed capacity — not the number of people actively working, available, or seeking security roles. Whether a given employer experiences a shortage depends on:

• How many licence holders in a given area are actively seeking or undertaking security work
• Where they are based and which roles they are qualified and willing to fill
• Pay, hours, and working conditions on offer
• Local employer demand and the number of open vacancies

A national decline in licence holders is a signal worth monitoring, not proof of a shortage at every company or in every region.

Evidence from employers or recruiters — applications per vacancy, time to hire, unfilled shifts — would be needed to establish that link with any confidence.

Our interview on the future of security supervision offers a supervisor’s perspective on pay, staff retention and supporting security teams.

What employers should monitor now

A national figure is only useful if it prompts a local check. Employers and operational managers can act on this now by tracking:

• Licence expiry dates — identify upcoming renewals and discuss employees’ intentions early, rather than discovering a gap at the point of expiry.
• Recruitment results — track suitable applicants, offers accepted, and time to fill vacancies against previous periods.
• Retention — record departures and the reasons given, so patterns become visible rather than anecdotal.
• Training access — establish whether cost, scheduling, or course availability is delaying applications or renewals among your own staff. For a separate example of longer-term staff development, our coverage of G4S’s apprenticeship training campaign looks at employer-led career progression.
• Operational coverage — monitor overtime and unfilled shifts alongside headcount, since these often show pressure before the licensing figures do.

These measures will not resolve the national question, but they will tell an individual business how exposed it is to a smaller licensed population, which is the more immediately useful thing to know.

A reversal to monitor

The headline finding stands: 447,558 active licence holders in September 2026, down 10,136 on the year before, alongside a substantial fall in first-time paid applications.

Several plausible contributing factors have been identified — new training requirements, the withdrawal of the £20 licence-fee rebate, cost pressures on individual applicants, and reported experiences of workplace violence.

The available sources document these pressures, but they do not establish how much, if any, of the decline each one caused. The evidence currently available supports treating this as a reversal that warrants further investigation rather than a fully explained one.

At the same time, the longer view provides useful proportion. The licensed population remains considerably larger than it was in 2021, and this is the first fall after four years of growth, not a multi-year decline.

Future releases should help resolve three things: whether the decline continues into 2027, which licence categories and regions are most affected, and whether employers report a corresponding change in staffing availability on the ground.

The central question remains open: is this a temporary fluctuation, or the start of a sustained reduction in the UK’s active SIA licence-holder population?

The Security Industry Authority (SIA) publishes monthly statistics on licence holders and licence applications, including breakdowns by age, gender, licence type and location. These freely available reports allow readers to explore the figures and track changes over time. View the latest licence-holder statistics and application data on GOV.UK