Britain's safety tech sector grew up in public

Revenue up 13%. Headcount flat. Investment down 43%. Those three lines describe the same British industry in the same year, and read together they say something the sector’s own promotional language does not.

The figures come from the UK Safety Tech Sector: 2025 analysis, prepared for DSIT by Perspective Economics and published on 10 April 2026. It identified 145 dedicated safety tech organisations in the UK, generating £704 million of revenue in 2024, up from £623 million in 2023, and about £226 million in 2019. That is a compound annual growth rate near 26% across five years, which almost no other British technology segment can claim.

Employment is roughly 3,900 full-time equivalents, with no significant net change on the previous year, and the report puts that at a quarter of a global safety tech workforce of about 15,700. Investment tells the opposite story: £24 million raised across 15 deals in 2024, down from £42 million across 18 deals the year before.

Revenue without hiring, growth without funding

A sector whose revenue rises 13% while its headcount stays still is either becoming more efficient or selling something that scales without people. In content classification and age assurance, both are plausible, and the second is more likely. Detection is a model, an inference bill and a support contract. Doubling the number of images checked does not double the staff.

The investment fall is harder to read charitably. Deal values move on small numbers and one absent round can produce a 43% drop, so the figure should not carry more weight than a fifteen-deal sample can bear. What it does suggest is that the growth is being funded by customers rather than by investors, which is a healthier position operationally and a weaker one strategically. Companies that grow on revenue grow at the speed their customers buy.

The size distribution supports that reading. Of the 145 firms, 78 are micro businesses of ten people or fewer and 37 are small. There are 27 medium-sized firms, up from 15 in 2023, and three large ones, up from one. A cohort is graduating, slowly, out of a very long tail.

The demand curve has a commencement date

Safety tech is unusual among British technology sectors in that the shape of its market is set by a statute with published dates.

The Online Safety Act passed into law on 26 October 2023. The government’s explainer records that in-scope services had to complete illegal content risk assessments by 16 March 2025, with Ofcom able to enforce from 17 March. Child safety duties followed, with children’s access assessments due by 16 April 2025 and risk assessments by 24 July 2025. Section 81, covering age checks, came into force on 17 January 2025. The maximum penalty is whichever is the greater of £18 million and a tenth of qualifying worldwide revenue.

That is a procurement calendar dressed as a regulatory one. The sector report says as much, noting that implementation and enforcement of the Act continue to drive demand, with Ofcom’s enforcement programme now assessing compliance.

What that means for the firms

Regulation-driven demand is the best and worst thing that can happen to a young industry.

The best, because it converts a discretionary purchase into a compliance obligation with a named accountable officer and a deadline. Nothing else moves an enterprise buying cycle so reliably. It also gives British suppliers a genuine head start, since the duties bite in the UK first and the expertise accumulates here. Fifty-nine per cent of providers already export, and 47% have a presence outside London and the South East, which is an unusual distribution for a British technology sector and worth more attention than it gets.

The worst, because demand created by a deadline arrives in a wave and then flattens. A platform that bought age assurance in 2025 to meet a duty is not buying it again in 2026. Renewal pricing on a compliance product is under pressure from the moment the risk of not having it drops from existential to routine, and the buyer knows the supplier’s alternative customers are also finite. Three large firms and 78 micro ones is a market shape that consolidates when growth slows.

The second-order risk is definitional. Compliance products get bought to satisfy a regulator, not to solve a problem, and the two diverge quietly. A service can pass an audit on the strength of a tool that catches the categories the guidance names while missing what actually harms its users. Nobody in that chain has an incentive to discover the difference.

Why this counts as a public sector story

Safety tech is usually filed under consumer technology, because the products end up inside social networks and games. That misreads where the money comes from.

The sector’s revenue is a function of a regulatory regime designed by a department, enforced by a regulator, and paid for by companies that would mostly rather not buy it. That is the same structure as environmental compliance or financial reporting software, and it produces the same industry dynamics: high certainty of demand, low customer enthusiasm, and pricing anchored to the cost of the penalty rather than to the value of the outcome.

It also means the sector’s fate is a policy variable. A change in Ofcom’s guidance on what constitutes highly effective age assurance moves more revenue than any product launch. Firms in this position are, in effect, running a regulatory forecasting function alongside a technology business, and the small ones cannot afford to.

The public interest test is different again, and it is the one nobody in the market has an incentive to apply. A tool that reduces reported harm is doing the job. A tool that reduces reported harm by removing more content than it needs to is passing an audit while shifting a cost onto users who never appear in the statistics.

What would settle it

The sector’s own metrics will not answer the question, because revenue growth is consistent with both the healthy and the unhealthy version of this story.

Three things would. Whether medium-sized firms keep converting from small ones once the initial compliance wave passes, since that cohort grew from 15 to 27 in two years while enforcement was ramping. Whether headcount starts moving, because a genuinely expanding market eventually needs people. And whether export share rises, since a domestic sector selling into a UK statutory deadline is a different business from one selling globally on capability.

Ofcom publishes enforcement activity and DSIT commissions the sector analysis annually. Both will be countable in a year, by anyone, without a briefing.

The regulatory strand of this sits with addressing the challenge of online harms, and the industrial policy question belongs with the rest of the public sector technology coverage, because a market conjured by statute is a public sector story whoever ends up paying the invoice.