I walked into a residential building in Cape Town earlier this year. Not a SecaTeq site. Premium development. R3 million apartments. Concierge desk. Designer finishes. The kind of place where every detail has been considered by an architect and approved by a developer who cares about quality. Then I looked at the security. A guard sitting behind a desk with a hardcover A4 notebook. Visitors sign in with a pen. No ID verification. No digital record. No way for the building manager to know who entered the building at 2am last Tuesday without physically driving to site and flipping through pages of illegible handwriting. This is not an exception. This is the norm. And it is a ticking liability for every body corporate, managing agent, and property developer in South Africa.
The Numbers Tell a Stark Story
South Africa’s private security industry is among the largest in the world. PSIRA’s 2024 annual report records more than 14,000 registered security businesses and over 2.7 million registered security officers. That second number is the one most often quoted, and it is the one most often misunderstood. Registered is not employed. The number actually working in the industry is 564,540. Set against a South African Police Service fixed establishment of 182,126, that still means roughly three actively employed private security officers for every person on the police payroll. The industry is not small. It is not niche. It is a critical pillar of how South Africans protect their homes, their businesses, and their communities.
A study by Tladi Paul Maja and Kris Pillay, published in the Southern African Journal of Security in March 2026, found what many of us in the industry already know: there are significant gaps in digital competency among security personnel, and the regulatory framework under the Private Security Industry Regulation Act 56 of 2001 is outdated when it comes to governing AI and emerging technologies in private security. The study called for inclusive training programmes and updated legislation. I agree. But we cannot wait for regulation to catch up. The buildings we protect need better systems today. (Maja & Pillay, Southern African Journal of Security, 2026. DOI 10.25159/3005-4222/20848.)
The Paper Problem Is a Data Problem
When a guard writes an incident in a paper OB book, that information dies on the page. It cannot be searched. It cannot be cross-referenced. It cannot be analysed for patterns. It cannot be shared with the managing agent in real time. It cannot be used to identify that the same vagrant has been removed from the same stairwell fourteen times this quarter, suggesting a structural access control failure rather than a policing issue.
When a patrol is tracked with a clipboard and a signature sheet, there is no way to verify that the guard actually walked the route. There is no GPS trail. There is no timestamp that cannot be falsified. There is no data that tells you which corridors are patrolled most frequently and which are consistently skipped.
This is not a technology problem. It is a governance problem. Body corporates and managing agents are making decisions about security budgets, guard deployments, and risk mitigation strategies based on no data at all. They are flying blind. And when something goes wrong — a break-in, an assault in a parking garage, a liability claim — the only evidence is a smudged entry in a notebook that may or may not hold up in court.
We stopped asking “did the guard do his patrol?” and started asking “what did the data from the patrol tell us about the building’s risk profile?” That shift changes everything.
— Rujhio Phillips, Group Chief Operations Officer, SecaTeq GroupWhat the Alternative Looks Like
At SecaTeq, we made a decision early on that technology would not be an add-on. It would be the foundation. Every guard in our operation clocks in digitally — verified, timestamped, and GPS-tagged to confirm they are physically at the site they are assigned to. No shared PINs. No paper registers. No workarounds. If a guard is not on site, we know within minutes. Not hours. Not the next morning when someone checks the register.
Every patrol is tracked via GPS in real time. The route, the duration, the checkpoints hit and missed. This data identifies patterns: which guards consistently complete patrols early, which corridors have a higher incident frequency on specific days, and which sites are trending toward a risk threshold that warrants intervention before something happens.
Every incident is logged digitally with photographs, GPS coordinates, and automatic categorisation. The managing agent sees it in real time. The body corporate chairperson gets a monthly report that is not a Word document typed by an administrator — it is a data-driven intelligence briefing drawn from live operational data.
What the AI Policy Withdrawal Actually Teaches Us
In April 2026, the Department of Communications and Digital Technologies gazetted South Africa’s Draft National AI Policy for public comment. Within weeks the Minister withdrew it. News24 had found that academic sources in the reference list did not exist. An internal review confirmed the citations had been generated by AI and published without anyone checking them. The policy was formally retracted by gazette in June, two officials were suspended for failing to declare they had used AI, and South Africa still has no national AI policy.
It would be easy to score a cheap point off that. I would rather draw the honest lesson, because it is the same one this article is about. The department did not fail because it used AI. It failed because nobody verified the output before it went out under a national letterhead. The Minister, Solly Malatsi, said the episode showed why vigilant human oversight over the use of artificial intelligence is critical. He is right.
That is exactly the risk in our industry. AI that categorises an incident, flags a patrol anomaly or drafts a client report is only as good as the person who checks it before it reaches a trustee. A security company that lets AI write its reports unchecked is one invented incident away from the same headline. Every report we issue is read by a human being who is accountable for it before a client ever sees it. That is not a feature we advertise. It is the minimum standard.
The regulatory vacuum does not remove the obligation. POPIA already governs how we handle personal information, biometric data and employee monitoring, and it applies today, policy or no policy. At SecaTeq we have built our systems with POPIA compliance at the core. Consent is explicit and recorded. Data is encrypted. Access is role-based. The companies that build verification into their process now will be ready whenever the policy finally lands. The ones waiting to be told what to do are already behind.
What This Means for Managing Agents and Developers
If you manage a portfolio of residential buildings in Cape Town, ask yourself three questions:
First: Can you tell me, right now, whether every guard assigned to your buildings last night actually showed up, on time, at the correct site? If your answer involves phone calls, WhatsApp messages, or checking with a site manager tomorrow morning, your system has failed.
Second: If there was a security incident at one of your buildings at 3am last Saturday, can you produce a digital incident report with photographs, GPS location, guard response timeline, and AI-categorised severity within 60 seconds? If not, ask yourself what you would present to the body corporate trustees. Or to an insurance assessor. Or to the police.
Third: Do you know which of your buildings has the highest incident frequency per unit? Which has the slowest guard response time? Which has deteriorating patrol compliance? If you cannot answer these questions with data, you are not managing security. You are hoping for the best.
The gap between what technology can deliver and what most security providers actually offer is the widest it has ever been. That gap is where risk lives.
— Rujhio PhillipsThe Shift Is Already Happening
The buildings that get this right will attract better tenants, command higher premiums, and carry lower insurance risk profiles. The managing agents who demand technology-backed security from their service providers will differentiate themselves in a market where most agents offer the same thing: a guard, a radio, and a notebook.
We are not suggesting that technology replaces guards. It does not. A camera cannot detain a trespasser. An algorithm cannot comfort a resident after a break-in. Security will always be a people business. But the tools those people use, the data those tools generate, and the intelligence that data produces — that is what separates a security company from a security partner.
At SecaTeq, we operate across Cape Town, from one of the largest residential complexes on the Atlantic Seaboard to a heritage golf course. Every site runs on the same technology platform. Every guard carries the same digital toolkit. Every client receives the same standard of reporting and transparency. We are not the largest security company in South Africa. But we believe we are building the most intelligent one.
The OB book served its purpose. It is time to close it.
See the Technology in Action
Book a free site assessment. No obligation. We will show you what your building’s security data looks like when it is actually captured, analysed and reported properly.
Rujhio Phillips
Rujhio leads daily operations across SecaTeq Group’s security, facilities management, and hygiene divisions. He oversees the deployment of digital workforce management technology across Cape Town and is responsible for client relations, staff performance, and regulatory compliance. Based in Cape Town, he is committed to raising the technology standard of the private security industry in South Africa.