Artificial Intelligence is rapidly moving from experimentation to operational reality.
Across manufacturing, AI is already helping organisations optimise production schedules, improve quality control, predict equipment failures, streamline supply chains, and support engineering and customer service teams.
The question is no longer whether AI will impact manufacturing.
The question is how organisations can use AI safely, responsibly and effectively.
That is where the EU AI Act comes in.
What Is the EU AI Act?
The EU AI Act is perhaps the most significant AI regulatory framework introduced to date and is expected to influence AI governance well beyond Europe.
Its purpose is not to stop organisations using AI. Quite the opposite.
The intention is to create a framework that encourages innovation whilst ensuring AI is deployed responsibly, transparently and safely.
The Act takes a risk-based approach. The higher the potential risk to people, organisations or society, the greater the level of oversight and governance required.
For most manufacturers and other organisations, this means understanding where AI is being used, how it is being used, and whether appropriate controls are in place.
The EU AI Act is essentially asking five questions:
- Do you know where AI is being used?
- Can you trust the data feeding it?
- Can you explain what it is doing?
- Does somebody own the outcome?
- Is the level of control proportionate to the level of risk?
Interestingly, these are not actually AI questions at all. They are the same leadership, governance, accountability and operational management questions that organisations should already be asking of every critical business capability.
The good news is that most manufacturers are not developing highly complex AI models themselves. They are typically deploying AI capabilities embedded within software platforms, machinery, industrial systems or operational workflows.
Why Has the Compliance Deadline Been Delayed?
One of the more interesting developments has been the decision to delay certain aspects of the compliance timetable, from August 2026 to later next year.
At first glance, some organisations may assume this means regulators are backing away from AI governance. That is not what has happened.
The technology has moved faster than the supporting operating model. In many respects, this is exactly the challenge organisations themselves are facing. AI capability is advancing rapidly, but governance, accountability, data quality, skills and operational processes often struggle to keep pace.
Regulators recognised that businesses cannot realistically comply with requirements when some of the supporting standards, guidance documents, testing procedures and certification mechanisms are still being developed.
In other words, the delay is less about relaxing the rules and more about ensuring organisations have a realistic opportunity to comply with them.
In my view, this should be viewed as a positive step rather than a retreat.
So, What Should Manufacturers Be Doing Now?
Whilst some compliance dates may have moved, manufacturers should not view this as a reason to pause. Instead, it provides valuable time to prepare.
There are four practical areas worth focusing on.
1. Understand Where AI Is Being Used
Many organisations are already using AI without fully realising it.
AI capabilities may exist within:
• Production planning systems
• Scheduling tools
• Predictive maintenance platforms
• Quality assurance systems
• Customer support applications
• Engineering and design software
The first step is simply visibility. After all, you cannot govern what you cannot see.
2. Review Data Quality
AI relies heavily on data.
Poor-quality data often produces poor-quality outcomes.
Manufacturers should ensure that data used within AI-enabled processes is accurate, consistent and appropriately governed.
The old principle of “poor data in, poor data out” (or words to that effect) still applies.
3. Establish Clear Accountability
One of the most common challenges I encounter across organisations is that nobody truly owns an outcome, end-to-end. Responsibility is often fragmented across departments, suppliers, technologies and processes. If nobody owns the customer outcome end-to-end, then nobody really owns the outcome at all.
AI does not solve this problem. In fact, it usually exposes it.
Clear ownership and accountability become even more important as AI adoption increases.
4. Strengthen Governance
Organisations should begin developing governance structures that answer simple questions, such as:
• Who approves AI use cases?
• How are decisions monitored?
• How are risks assessed?
• How are outputs reviewed?
• How do we demonstrate compliance if required?
The organisations that establish these disciplines early will find future compliance significantly easier.
The Bigger Challenge Isn’t AI
In my experience, the biggest obstacle facing most organisations is not the technology itself. After all, AI didn’t create disconnected departments, fragmented workflows, poor data quality or unclear accountability. Those issues already existed. AI simply shines a brighter light on them.
So, the biggest obstacle is the operational environment into which that technology is being introduced.
Many manufacturers still operate across:
• Organisational silos
• Disconnected systems
• Fragmented workflows
• Multiple data sources
• Inconsistent processes
AI did not create these issues. But it makes them much more visible.
This is why discussions about AI readiness should really begin with service and operational readiness.
Before organisations can fully exploit AI, they need confidence that their people, processes, technology, data and knowledge are working together as one connected service system.
Without those foundations, AI risks scaling inefficiency.
With them, AI can become a powerful accelerator of productivity, innovation and customer value. Herein lies the significant opportunity for use of AI in manufacturing.
Manufacturing Is Becoming More Service-Led
There is another important trend running alongside AI.
Many manufacturers are evolving beyond simply selling products. Increasingly, they are selling ‘outcomes’. This is one of the driving forces behind the continued shift towards Equipment-as-a-Service, Outcome-as-a-Service and wider Everything-as-a-Service (XaaS) business models.
We see this through:
• Predictive maintenance services
• Remote monitoring
• Performance guarantees
• Subscription-based offerings
• Equipment-as-a-Service models
• Outcome-based commercial agreements
Customers are becoming less interested in the asset (product/equipment/machine) itself and more interested in what that asset enables them to achieve.
That shift places even greater importance on data, service delivery, operational visibility and customer outcomes.
AI has the potential to accelerate this transformation significantly.
Five Questions Every Manufacturing Leader Should Therefore Be Asking
As AI adoption continues to grow, every manufacturing organisation should be asking:
- Do we know where AI is currently being used across our business?
- Can we explain how AI-supported decisions are being made?
- Is the data feeding those systems trustworthy?
- Who owns the customer outcome end-to-end?
- Are we still primarily selling products, or are we increasingly delivering outcomes?
Final Thoughts
The delay to elements of the EU AI Act should not be viewed as a reason to slow down.
Instead, it should be viewed as an opportunity to strengthen the foundations that successful AI adoption depends upon.
The manufacturers that thrive over the next decade will not necessarily be those with the most AI. They will be those with the most mature service operating models, capable of turning AI, data and connected products into measurable customer outcomes.
AI does not replace service. AI executes through service.
The organisations that understand this distinction earliest will be the ones best positioned to create new value, launch new advanced services, strengthen customer relationships and compete successfully in an increasingly outcome-driven economy.
After all, customers rarely buy technology. They buy the outcome that technology enables.
Johann Diaz
Service Revolution Academy
www.Service-Revolution.com
Connect with Johann on Linkedin
Further Reading
AI Act high-risk system guidance
AI in Manufacturing March 2026: Key Launches, Strategy Shifts and What They Actually Mean
Digital Twin in Manufacturing 2026: The Plain-English Guide for Factories
What Is Edge AI in Manufacturing and Why It Matters in 2026
Big Tech Has Found the Factory Floor
FAQ
What does the EU AI Act mean for manufacturing leaders?
The EU AI Act means manufacturing leaders need to understand where AI is being used, what data feeds those systems, who owns the outcomes, and whether the right governance controls are in place.
Do manufacturers need to stop using AI because of the EU AI Act?
No. The EU AI Act is not designed to stop AI adoption. Its purpose is to support safer, more transparent and more responsible AI deployment, especially where AI could affect people, safety, operations or critical business outcomes.
Why is AI governance important in manufacturing?
AI governance is important because manufacturing AI often sits inside production planning, predictive maintenance, quality control, customer support, engineering software and industrial workflows. Without governance, organisations may struggle to explain decisions, manage risk or prove accountability.
What should manufacturers do first to prepare for the EU AI Act?
Manufacturers should start by mapping where AI is already being used across the business. This includes AI built into machinery, production systems, quality platforms, scheduling tools, engineering software and customer-facing workflows.
Is the EU AI Act only relevant to companies developing AI models?
No. Many manufacturers may not develop AI models themselves, but they often deploy AI through third-party software, machinery, industrial platforms or operational systems. That still creates a need for visibility, accountability and governance.
