Artificial intelligence (AI) has quietly become an integral part of everyday operations in many factories. It is being used to optimise production schedules, predict equipment failures, identify quality defects, and support engineering teams during product development.
When the EU AI Act is mentioned, many business leaders immediately think about generative AI, chatbots, or systems like ChatGPT. While these technologies continue to attract attention, they are not necessarily where the most significant regulatory impact for manufacturers will arise.
For manufacturers, the more important question is whether AI influences how products behave, how machinery operates, or how safety-related decisions are made.
Not all AI systems create the same regulatory exposure.
Consider two examples.
In the first case, an AI model analyses vibration data from industrial equipment and predicts that a bearing is likely to fail within the next few weeks. The system generates an alert and recommends maintenance. An engineer reviews the information and decides what action should be taken.
In the second case, an AI-powered vision system monitors a robotic cell. When a person enters a hazardous area, the AI detects the situation and automatically triggers a safe stop or reduces the robot's speed.
Both solutions rely on advanced AI technologies.
However, their role within the manufacturing process is fundamentally different. The first system supports human decision-making. The second system directly contributes to a safety-related function. Whether this ultimately leads to a high-risk classification depends on the applicable product legislation and conformity assessment route.
As AI becomes more closely connected to physical machinery and product safety, the regulatory assessment becomes more important and potentially more complex. This distinction is likely to become one of the most important topics manufacturing companies will need to understand over the coming years.
The AI Act is often discussed as a technology regulation. For manufacturers, however, it should be viewed equally as a product safety and engineering challenge.
Industrial companies already operate in a highly regulated environment. Product conformity, quality management, risk assessments, technical documentation, validation activities, and CE marking are familiar concepts for engineering and quality teams.
The introduction of AI does not replace these processes. Instead, it adds a new dimension to them.
As AI becomes embedded in products and machinery, manufacturers increasingly need to address AI-related risks alongside traditional product safety requirements. This turns AI compliance into a cross-functional effort involving engineering, product safety, quality assurance, cybersecurity, and compliance teams.
Given the length of product development and certification cycles in manufacturing, early preparation can be critical to avoid regulatory requirements becoming a late-stage constraint.
The good news is that most manufacturers do not need to reinvent their operating model. In many cases, existing product safety, quality, and risk management processes provide a strong foundation for future AI-related obligations.
AI is increasingly embedded in internally developed applications, machinery, software platforms, industrial automation solutions, vision systems, and supplier-provided equipment. In many organisations, a surprisingly difficult question is simply:
The first step is therefore understanding where AI is present across engineering, production, maintenance, quality control, and supplier ecosystems.
The next challenge is determining which systems merely support human decisions and which have a direct impact on machinery behaviour, product performance, or safety outcomes.
That distinction will increasingly shape how manufacturers assess AI-related risks, compliance requirements, and evidence across the lifecycle of industrial systems.
Manufacturers do not need to start from scratch. We help organisations identify and classify AI use cases across products, machinery, OT environments, and production processes, assess their potential regulatory exposure, and determine whether product safety and conformity assessment requirements may apply.
By connecting AI-related obligations with existing quality management systems, technical documentation, risk management, and lifecycle processes, we help companies establish a practical governance model that enables responsible AI adoption while maintaining product safety, compliance, and operational efficiency.
The key challenge for manufacturers is extending existing product safety, quality, and compliance frameworks to cover AI-enabled products and systems.
Success starts with understanding where AI is used and how it affects products, machinery, and production processes. Companies that establish this visibility early will be better positioned not only to address regulatory requirements and future conformity assessments, but also to scale AI with greater confidence across their operations and products.