Atomicwork passes its first ISO 42001 surveillance audit
Any company can announce a certification. Far fewer can show that, a year later, the system behind it still holds up to an outside auditor with no reason to be generous. That is the real test of an AI management system.
Atomicwork has now passed that test. Our first ISO/IEC 42001:2023 surveillance audit, conducted by our certification body INTERCERT in October 2026, confirmed that our AI management system is still operating and improving. We first earned the certification in September 2025.
ISO/IEC 42001:2023 is the world’s first international standard for AI management systems. It sets the bar for how an organization governs the AI it builds and operates: how risk is assessed, how decisions are documented, how transparency and accountability are enforced, and how all of it is reviewed over time.
What is an ISO 42001 surveillance audit?
A surveillance audit is a follow-up audit, usually held every year, in which an independent certification body checks that a certified management system is still operating and improving.
A first-time certification captures a moment. You stand up the controls, assemble the evidence, and demonstrate the system on the day of the audit. A surveillance audit is different. It arrives a year on and asks a harder question: is this still how you actually operate, or was it a project that ended once the badge was framed on the wall?
Clearing it means our AI governance is not a static document set. The risk assessments are current. The impact assessments are being run. The controls are maintained quarter after quarter, the decisions are documented as they are made, and the evidence is there when an independent assessor comes looking.
Standards are easy to win and hard to keep. We kept this one.
Why ISO 42001 matters for an AI Workforce platform
Atomicwork builds AI-native IT and enterprise service management software. Our AI Workforce is made up of AI Coworkers that take on real roles inside enterprise IT and operations, carry their own scoped access, and act on behalf of real people. Customers give these systems real authority over their environments, so they are right to ask how the company behind them governs its own AI.
ISO 42001 is how we answer that question with evidence rather than assurances. The same discipline our customers expect from an AI platform, safety, transparency, accountability, and human oversight, is the discipline an ISO 42001 audit measures. Passing the surveillance audit is external validation that the way we build and run AI is held to a standard, not to a slide.
Where Atomicwork stands on ISO 42001
ISO 42001 was only published in December 2023, and independent certification is still an early signal in enterprise AI. Atomicwork was among the first to certify and has now passed a surveillance audit. In a market where “responsible AI” is said often and proven rarely, maintaining an independently audited AI Management System puts Atomicwork ahead of most of the field, not alongside it.
What comes next for Atomicwork’s AI governance?
Responsible AI is not a destination, and a certification is not a finish line. The standard itself assumes continuous improvement, and so do we. We will keep running our risk and impact assessments, keep tightening our controls as our AI Workforce grows, and keep inviting independent scrutiny because it keeps us honest. The goal is simple: as the number of AI systems we operate increases, the governance around them should get stronger, not thinner.
Our thanks to every team across Atomicwork whose everyday discipline made this outcome possible, and to INTERCERT for a rigorous and professional audit.
Talk to us if you want to know more about how we govern AI and embed trust and compliance into our platform.
