The pros and cons of AI in quality management, laid out honestly
Eight out of ten organizations in the QHSE sector are still in the experimental phase when it comes to AI. Only a fraction have an integrated approach. (Source: National QHSE Survey).
At the same time, AI is on almost every agenda. This tension between interest and hesitation is easy to explain. AI can save quality managers a lot of time, especially with administrative tasks, but its use also carries risks. Think of privacy, reliability, and dependency. In this article, we list the pros and cons for you.
Where the hesitation comes from
In practice, AI is often used for relatively simple processes. For example, writing texts, summarizing documents, and/or brainstorming ideas with tools like Microsoft Copilot, ChatGPT, Google Gemini or Claude. Organizations that are truly daring enough to deploy AI more broadly within their QHSE processes are still the exception at the moment.
A cautious approach is easy to explain. Quality management is all about reliable information, verifiable processes, and risk control. New technology only earns a permanent place once organizations can trust that it aligns with the requirements of quality management.
Less administration, more substance
For quality managers, time savings are the biggest advantage of AI. A lot of time in a workday is spent on administrative tasks. Think of drafting documentation, processing registrations, and gathering information. The National QHSE Survey shows that organizations see efficiency improvement (82%) as the most important benefit of AI, followed by faster problem-solving (69%) and higher productivity (53%). Quality managers also indicate that they want to spend more time on work that truly adds value.
In the survey, QHSE managers often describe AI as a sparring partner. Someone who helps with drafting documentation, looking up laws and regulations, and translating standard requirements into plain language. Think of a first draft for a procedure, a work instruction, or an audit plan. AI does not take over the work, but helps you reach a usable first version faster.
Time is also saved on the kind of data migration that nobody enjoys. A set of forms in Word or PDF that is automatically converted into digital forms. Screenshots or drawings of flowcharts that become clickable process flows in a single click. Previously, this required manual labor or an expensive integration that you had to maintain.
Routine tasks take less time, creating more room for work where the quality manager can make a real difference. For example, analyzing risks, guiding improvement projects, advising colleagues, and strengthening the quality culture. This is where the true strength of the profession lies. AI offers support, but the substantive assessment and final responsibility remain with the quality manager.
What you need to take seriously
AI offers many possibilities, but it also carries risks that you, as a quality manager, cannot ignore.
1. Privacy and intellectual property
Privacy and intellectual property are at the top of the list of concerns for QHSE managers. It becomes a risk when an audit report, complaint log, or other document containing company-sensitive information accidentally ends up in a public chatbot. Organizations that operate according to ISO 27001 must, in any case, be fully aware of where data is processed and who has access to it. Responsible AI use requires clear agreements on what information may and may not be shared.
2. Reliability of the output
AI sometimes provides convincing answers that are factually incorrect or outdated. Think of a misinterpretation of a standard or a reference to obsolete legislation. Procedures and work instructions form the basis for audits and daily operations. Such errors can therefore have major consequences. A review by an expert remains essential before information is adopted or shared.
3. Dependency
Dependency is the third risk. Employees may start to rely too heavily on AI, causing them to think less critically themselves. AI can provide a good initial answer, but it does not understand an organization's context the way an experienced quality manager does. If suggestions are adopted without verification, errors can end up in processes or documentation unnoticed. AI is a tool, not a replacement for professional expertise.
4. Sustainability
Sustainability deserves a conscious trade-off. AI requires significant computing power and, consequently, a lot of energy. Many QHSE managers are responsible for their organization's sustainability goals, which requires a deliberate approach. Use AI consciously and purposefully to keep efficiency and sustainability in balance.
5. Costs
The costs of self-developed solutions are often underestimated. A custom 'GPT' or internal app requires not only development but also maintenance, management, and validation. Gartner expects that by 2028, at least half of all AI projects will exceed their budget due to poor choices and a lack of operational knowledge. Organizations that build their own AI solutions often abandon the attempt due to costs, complexity, and technical debt. (Source: Gartner, Hype Cycle for Generative AI 2026)
Who is responsible when AI makes mistakes?
This question will come up sooner or later, especially at companies with certifications. The answer is simpler than it seems: The responsibility lies with the person using the tool. As Edward Niemeijer phrased it in the webinarThe QHSE manager of tomorrow, AI in practice: "If someone makes a mistake with a saw, it's not the saw's fault."
It gets more complicated with systems that independently draw conclusions that affect people. The childcare benefits scandal is the best-known example of this: a system produced an output that no one critically reviewed anymore. That is exactly the situation you want to avoid within QHSE, according to Edward Niemeijer from the AI expertise center.
In practical terms, this means you identify the points in your processes where a human must always review things before they are sent out. At a minimum, this applies to areas where the consequences are greatest, such as anything going to customers, certification bodies, or the shop floor.
Four steps to responsible AI use
Step 1. Get your documentation in order first.
AI only functions well if the foundation is solid. Fragmented information with multiple versions of documents will also lead to unreliable answers from AI. An up-to-date management system with validated processes, procedures, and documents forms the foundation for responsible AI use. Want to get started with AI? Begin by cleaning up and centralizing your management.
Step 2. Define which data can be processed where.
Public AI tools are not suitable for confidential company information. Therefore, clearly define which data employees may and may not use and establish clear agreements. This limits privacy risks and keeps you aligned with ISO 27001 requirements.
Step 3. Identify critical processes.
Determine for each process where an expert should review the output before it is used or shared. Think of AI as a digital colleague (you wouldn't pass on a colleague's work without looking at it either).
Step 4. Assign responsibility and evaluate.
Establish who is responsible for AI-generated output and include its use in internal audits and management reviews. This ensures AI truly becomes part of your management system rather than something happening on the side.
ISO 42001, the standard most QHSE managers don't know yet
It is striking that 79% of QHSE professionals are not yet familiar with ISO 42001. This is notable because this standard is the management system standard for AI and follows the same logic as ISO 9001 and ISO 27001.
For most organizations, certification is not yet on the agenda. However, the structure of the standard is useful as a framework, even without certification. The standard compels you to document where AI is being used, the associated risks, and who is overseeing it.
The low awareness of ISO 42001 is directly linked to the limited intention to certify. Combined with the low AI maturity of organizations, this indicates that AI management systems and associated certification are still a future issue for most organizations, rather than a current priority.
What does this mean for your organization?
AI can offer organizations significant benefits, provided the foundation is in order. An up-to-date management system, clear processes, and reliable documentation are essential starting points for responsible AI use. Only then can you leverage AI to perform tasks more smartly and efficiently.
ISO2HANDLE supports organizations with a platform where documentation, forms, and processes are managed centrally. Within this environment, AI functionalities are available, such as the AI Form Converter, the AI Processflow Converter, and the AI Form Translator. Integrations with Microsoft Copilot, Claude, and ChatGPT are also possible, allowing employees to work in their own environment without taking company information outside. Because ISO2HANDLE is ISO 27001 certified, the handling of information also aligns with organizations that prioritize privacy and information security.
AI is here to stay in quality management. The question is therefore not whether you should start using it, but how you can work with it responsibly.
Start small, always keep a human involved in important decisions, and ensure the foundation of your management system is truly in order. With a solid foundation, AI can grow into a valuable tool that supports the quality manager without taking over the reins.
Quality, Health, Safety and Environment?
