Building safety reform is quietly shifting the industry from point-in-time certification to continuous compliance, argues Gary Ng of viAct. Contractors that update only their paperwork, not their operating model, are at risk of falling behind, he warns
Generally speaking, if you ask a tier 1 contractor about whether their latest project complies with safety regulations, they would likely point to a certificate. But if you ask whether the project is still in compliance today, three months after that certificate was issued, you would find them to be less certain.
The difference between the certification and the actual state of compliance on the building site reveals the real story behind UK building safety reform, and it is bigger than any single Gateway.
Since the establishment of the Building Safety Regulator as an independent agency outside the Health & Safety Executive earlier this year, reform has moved consistently in one direction: from periodic checkpoints towards ongoing oversight. The government has been laying the groundwork for a single construction regulator with the aim of bringing coherence to a historically fragmented oversight system.
The shift in focus may be evident in the published performance statistics, although that shift may not be entirely reflected in the structure yet. The most recent quarterly figures show approval rates of 89% for new higher-risk buildings, 85% for external remediation, and 73% for internal works, with median approval times ranging from 22 to 35 weeks depending on category.
That’s useful operational progress, but it is still information reported checkpoint by checkpoint, category by category. It says very little about what’s happening on a site the week after a Gateway 2 approval is granted, or the week before Gateway 3.
“Contractors that treat each new consultation outcome as a fresh compliance project will keep finding themselves a step behind, reacting to whatever the regulator publishes next.”
The gap between certification and ongoing safety
The pattern that building safety reform has been pointing to is not restricted to a single Gateway. A pass at one stage says nothing about week six of construction. Certification is being regarded as a one-time event, whereas the safety risk it is meant to solve is ongoing.
Contractors that read this reform purely as “more paperwork, more Gateways” are dealing with last year’s issue. Those that establish real-time and continual visibility of site conditions, and not just audit-ready documentation, will absorb whatever comes out of the current consultation without a scramble. That’s a capability question, not a compliance question.
Continuous compliance in practice
The technology making this shift practical is AI-driven computer vision. The output from cameras and sensors already on site can be analysed continuously by AI, flagging deviations and risks as they emerge instead of waiting for the next scheduled review. Similar trends can be seen in other parts of the world that have made more progress down this path. Two examples illustrate what the operational payoff looks like once continuous monitoring replaces point-in-time checkpoints.
A contractor from Singapore involved in high-rise and infrastructure projects replaced traditional periodic safety checks with live, always-on AI monitoring of PPE compliance, equipment movement and access control. Instead of waiting for the next scheduled audit, the site team can now detect potential risks and eliminate them before it is too late. After adoption, the contractor reported a tenfold increase in its internal safety score across projects.
Similarly, a Hong Kong contractor delivering a large mixed-use development involving deep basement excavation in a dense urban corridor was relying on manual topographic surveys every four to six weeks. By the time deviation reports reached engineers, formwork had already been stripped, so corrections meant demolishing completed work.
The contractor switched to continuous LiDAR scanning and AI. When a threshold movement was flagged at an earth retaining wall, engineers were notified before the next shift began. Rework dropped by 60% and the team saved more than 3,200 hours previously spent on manual surveys.
These real-world cases serve as evidence of what AI-driven monitoring makes possible once contractors stop treating compliance as a single goal to be achieved, and start treating it as an ongoing responsibility that technology can help them sustain.
Building capability before reform requires it
None of this requires UK contractors to wait for a specific piece of legislation to land before acting. AI-based monitoring, using the technology that sites already have, makes that ongoing visibility achievable without adding headcount. The direction of every reform since the Building Safety Act – independent regulation, the push towards a single regulatory body – points the same way: oversight is moving from something granted once to something maintained continuously.
Circling back to the question this article opened with: ask a contractor if their site is compliant today, not at the point of certification, but right now, three months in, and the answer should be as confident as the one on the certificate itself. For most, it currently isn’t, and that gap is what the coming years of reform are quietly closing.
Contractors that treat each new consultation outcome as a fresh compliance project will keep finding themselves a step behind, reacting to whatever the regulator publishes next. Those that build genuine, AI-supported visibility into site conditions now are building something reform cannot make obsolete, because it’s the same underlying capability the regulator itself is edging towards.
That’s the real choice in front of tier 1 contractors: keep managing compliance Gateway by Gateway, or start treating safety as a continuous state the business is always demonstrating, whether or not anyone is currently inspecting.
Gary Ng is co-founder and CEO of viAct and has more than 10 years’ experience in implementing technological innovations in the construction industry.
