TL;DR

Safehire.ai has taken £500,000 from a private investor in its second seed raise, bringing total funding to £1.55m. The Gloucestershire company sells AI-assisted background screening into the education sector and intends to use the money to expand into larger enterprise deployments, build an insider risk capability, and move into other jurisdictions. The previous round, £1.05m, closed in April 2025.

What the platform does

Digital Risk Screening trawls publicly accessible sources — the ordinary web, the parts search engines do not index, and dark web material — then applies AI analysis before analysts trained in military intelligence verify the output. The company puts throughput above 10,000 searches daily, against 15 that one person could manage, and well below conventional cost, citing £1,500 as a typical professional vetting fee. An academy trust customer reported cutting roughly an hour of HR time from each candidate.

Simon Holden, who was a school bursar before founding a cybersecurity company, is based near Gloucester; his co-founder Sean Lumley, who started the cybersecurity platform CybSafe, works near Stroud. The two met during officer training at Sandhurst, which explains the analyst hiring pattern.

The claim, and the question under it

Holden’s framing is that conventional checks miss things: DBS and credit checks do not reliably surface online activity or behavioural signals, and the platform is sold as closing that visibility gap. That is a real gap in safeguarding terms. It is also a description of automated open-source intelligence gathering on job applicants, which sits in a different regulatory position from a DBS check — the latter is a statutory process with defined scope and a route of appeal.

The human-verification step is what the company points to as the control, and the analyst-review model is a more defensible design than a fully automated risk score. UK employers considering this category of tool will still need a position on what a candidate is told, what an adverse finding from web material actually evidences, and how someone contests it. Recent UK polling showing two thirds of Britons reject AI sifting of job applications suggests candidate acceptance is not assumed.

Looking forward

This is a small round by AI funding standards and would ordinarily stay a regional story. It is worth noting for two reasons: it is attributable South West deal flow of the kind that rarely reaches national coverage, and the move from education safeguarding into insider risk points a screening tool built for child protection at general corporate workforce monitoring.