The AI paradox in QHSE: between hesitation and hype

The toddler phase of the revolution

If we are to believe the media, AI will take over our jobs and control our factories tomorrow. However, the reality on the quality manager's shop floor is different. In fact, survey results show that in terms of AI maturity, the industry is not yet very advanced: the vast majority of organizations are in the ‘initiative phase’ (level 1) or ‘experimentation phase’ (level 2). Only a fraction of those surveyed dare to speak of an ‘integrated approach’.

The tools that are widely used are often limited to low-threshold solutions such as ChatGPT and Copilot for writing texts. The real depth – using AI for complex risk analysis or predictive compliance, for example – is still in the future for many. You mainly see a wait-and-see attitude: “We're waiting to see what happens”.

Fear of the ‘Black Box’

Why this reluctance? The report exposes a number of fundamental concerns that put the brakes on innovation. At the top of that list are privacy and intellectual property. Many QHSE professionals fear that confidential business information will be out in the open via public AI models or that the data will not stay within the organization's walls.

There is also a human, ethical dilemma at play. There is fear that employees will ‘stop thinking for themselves’ and blindly trust the output of an algorithm. The reliability of the information is also a tricky issue: is what AI is saying accurate, or is it a hallucination based on outdated legislation? Also notable is the concern about sustainability: the enormous energy consumption of AI systems conflicts with the very environmental goals that many QHSE managers are trying to achieve.

The promise: from administration to strategy

Yet, there is more than just skepticism. In fact, the ambition to truly integrate AI into business operations is there. QHSE professionals indicate that they see AI as a potentially powerful tool for efficiency improvement and risk management. People dream of having a ‘sparring partner’ to help formulate documentation and translate standards into practice.

This is not very surprising, as it aligns exactly with one of the great frustrations of the modern quality manager: the administrative burden. Quality managers want to do “less ‘admin’ work and more ‘value-adding’ work”. AI could be the key to freeing up (more) time for strategic consulting and culture change, rather than endlessly chasing checkmarks and document versions.

No AI without I (Information Management)

Here lies the crucial intersection where technology and quality management intersect. The fear of unreliable output and data breaches is justified as long as the basic information is not structured. After all, AI depends on the data it is fed with. Garbage in, garbage out – that's what it comes down to.

This is where WoodWing Scienta can become relevant to many organizations. Scienta ensures that an organization can run with AI by first teaching it to operate effectively through solid knowledge management. Scienta helps organizations provide centralized, structured, and accessible knowledge – exactly the necessary foundation on which any future AI application can lean.

If processes, protocols and work instructions are scattered across network drives or outdated, any AI tool will fail. The risk of ‘hallucinations’ decreases when the source – the quality management system – is validated and current because it operates on the concept that there is one source of truth. In the future, when AI is placed as a layer over such a centralized knowledge base, it suddenly ceases to be an unreliable oracle and instead provides rapid access to verified business knowledge.

The route forward: structure as a springboard

The results from the National QHSE Survey show that people remain the weakest link in information security. This is precisely why a closed, secure knowledge platform is essential. It prevents proliferation of unverified data in public AI tools - which are by definition unsafe.

For the QHSE manager of 2026, the message is clear: don't let the AI hype get your head spinning, but don't ignore development either. Start with the basics: make sure your processes are in place, that your knowledge is discoverable, and that your quality management ‘lives’ in a system that is both solid and future-proof. Then, you give AI the chance to exchange the status of ‘potential threat’ for that of ‘ultimate assistant’ that allows the QHSE manager to focus again on what he excels at: improving, connecting, and securing.

This article, based on the results of the Dutch National QHSE Survey, was published in print in Dutch magazine Kwaliteit in Bedrijf. Want to order the survey report (exclusively available in Dutch)? Visit the National QHSE Survey website for terms and conditions.