In brief
- 73% of manufacturing organisations report rapid AI adoption, with 90% ranking AI as a top security priority for 2026.
- Governance is advancing, but 53% have identified shadow AI usage and only 52% have a dedicated AI committee.
- Risk now forms through routine interactions, such as design files, production data and supplier contracts shared with AI tools.
- Small, everyday interactions can accumulate into meaningful IP exposure, so control must operate at the level of workflows and data.
Artificial intelligence is rapidly becoming embedded across manufacturing environments, from engineering and design to supply chain optimisation and operations.
What was once experimental is now being applied in day-to-day workflows, often driven by the need for speed, efficiency, and competitive advantage.
Recent research shows that 73% of manufacturing organisations report rapid AI adoption, with 90% ranking AI as a top security priority for 2026. The direction of travel is clear. AI is becoming a core part of how industrial businesses operate.
At the same time, the data suggests that control is still evolving.
The Hidden Complexity of AI in Industrial Environments
Manufacturing presents a distinct challenge when it comes to AI usage.
Unlike more centralised IT systems, AI is being used across a wide range of functions, including engineering teams working with proprietary designs, operations teams handling process data, and supply chain teams analysing sensitive commercial information.
This creates a highly distributed and data-rich environment.
While 66% of organisations report having governance policies in place, and 63% say they can detect sensitive data in AI tools, more than half (53%) have still identified shadow or unauthorised AI usage within their business.
This reflects how AI is being adopted in practice. Tools are often embedded into existing platforms, accessed via browsers, or introduced informally by teams seeking efficiency gains.
As a result, usage can quickly extend beyond what is formally approved or centrally visible.
From Systems Risk to Workflow Risk
In manufacturing, the nature of risk is also shifting.
Traditionally, security has focused on protecting systems, networks, and stored data. AI introduces a different model, where risk is created through interaction.
- A design file uploaded for analysis.
- A production dataset shared with an AI assistant.
- A supplier contract summarised using an external tool.
These actions are not unusual. In many cases, they are part of normal work.
However, they can involve highly sensitive intellectual property, operational data, or commercially critical information. This creates a gap between governance intent and operational reality.
Policies may define acceptable use, but they do not always influence behaviour at the point where data is being shared.
The IP Risk That’s Hardest to See
For manufacturing organisations, one of the most significant concerns is the potential exposure of intellectual property.
Unlike traditional data breaches, AI-related risks are often subtle. Information may be shared in small fragments, across multiple interactions, without triggering conventional alerts.
At the same time, once data is entered into an external AI system, control over how that data is stored, processed, or reused can become unclear.
This is particularly relevant in industries where competitive advantage is closely tied to proprietary designs, processes, or formulations.
The challenge is not just preventing large-scale leaks, but understanding how smaller, everyday interactions could accumulate into meaningful exposure.
Governance Is Advancing, But Not Yet Complete
The research suggests that manufacturing organisations are actively responding.
More than half have established governance policies, and many are beginning to formalise oversight through committees and structured approaches.
However, with only 52% reporting the presence of an AI committee, governance maturity remains uneven across the sector.
This is consistent with a broader transition phase.
AI adoption is being driven by business value, while governance structures are still catching up to the pace and nature of that adoption.
From Visibility to Control
What is becoming increasingly clear is that visibility alone is not sufficient.
Knowing which tools are being used is important, but it does not fully address how those tools are being used, what data is being shared, or whether policies are being followed in practice.
The focus is therefore shifting towards more granular control.
This includes understanding AI usage at the level of workflows and interactions, identifying where sensitive data is being exposed, and ensuring that safeguards operate in real time.
As CultureAI’s Chief Revenue Officer, Sam Soares, explains:
“In manufacturing, AI is being adopted quickly because the productivity gains are obvious. The challenge is that a lot of that usage is happening in engineering and operational workflows where sensitive data is constantly in play. The question is not whether AI is being used, but whether it’s being used in a way that protects IP and maintains control.”
A Defining Moment for Industrial AI
Manufacturing is entering a phase where AI is no longer optional, but neither is control.
The organisations that succeed will be those that can balance speed with oversight, enabling teams to use AI effectively while maintaining a clear understanding of how data is flowing through these systems.
This is not about slowing adoption. It is about ensuring that adoption is sustainable, secure, and aligned with long-term business value.
Because in manufacturing, the risk is not just data loss.
It is the potential loss of intellectual property, competitive advantage, and operational integrity.
📕 Explore the Full Findings
This blog only scratches the surface of the data and analysis behind these shifts.
👉 Read the full report: The State of Enterprise AI Usage: The Illusion of Control
📊 Start surfacing risks in your own organisation: AI Risk Assessment