The pressure is existential and very real. While four out of five CEOs are more optimistic about the ROI of their AI investments than they were a year ago, half told a World Economic Forum survey that their own jobs could be at risk if those investments fail to deliver results. AI is no longer in pilot territory. For CEOs, the challenge is now how to scale AI responsibly and profitably.
The challenge for Industrial CEOs: Ambition outpaces readiness
The industrial sector is at a critical inflection point. The global AI in manufacturing market is set to grow from $34bn in 2025 to $155bn by 2030. Yet an international survey conducted in 2025 for TATA/AWS found only 21 percent of manufacturers are fully prepared for AI adoption.
The stumbling blocks?
- Data readiness: Production data is trapped in legacy OT systems with no extraction or integration path.
- IT/OT silos: Data from CISCO shows only 20 percent of organizations have a fully collaborative IT/OT security posture.
- Skills mismatch: 60 percent of organizations cite capabilities gaps – not headcount – as the core workforce problem.
- Lack of governance: AI is deployed without policy, oversight, or clear accountability frameworks.
- Pilot paralysis: Strong proof of concept results never reach production.
These are the five hurdles industrial CEOs must address to scale AI successfully. And with TATA/AWS finding that 75 percent of manufacturers expect AI to be a top-three driver of operating margin this year, the time to act is now.
In the race to realize AI's potential, speed matters. CEOs who take decisive action today are better positioned to capture long-term value, while those who wait may find themselves struggling to catch up.
#1. Data readiness
Data is the non-negotiable foundation for AI success. Gartner has found that the single most common reason AI projects fail is data quality. For manufacturers grappling with legacy systems and an IT/OT disconnect, establishing a unified data layer that connects OT with IT systems is the foremost priority.
Edge compute capacity is an increasingly important facet of data readiness. CISCO’s 2026 State of Industrial AI report found 97 percent of industrial leaders expect AI workloads to significantly increase connectivity requirements. While 96 percent of firms see wireless reliability critical to industrial AI, preprocessing of data in local industrial systems before it is sent to the cloud for further analysis helps reduce the connectivity burden.
Modernization offers another step toward data integration. That’s why COPA-DATA’s zenon software platform supports a vast range of communication protocols and standards including support for legacy protocols. zenon becomes the secure bridge between legacy data and systems and modern data pipelines and platforms. Data silos can be eliminated, fostering a holistic view of operations, and enabling seamless interaction among all components of the industrial ecosystem for enhanced efficiency and collaboration.
#2. IT/OT silos
There’s a critical gap in the convergence of IT and OT systems: security. The Dragos OT Cybersecurity review of 2025 found 75 percent of OT attacks begin as IT breaches. Third-party vendors and contractors continue to be a major weak point, with some organizations unaware of all remote connections into their OT networks. Many OT environments still rely on outdated remote access policies, instead of role-based, monitored, and segmented access controls. And, of the vulnerabilities identified, 70 percent were deep within OT networks, making them difficult to patch without disrupting operations.
Furthermore, Dragos found that many organizations continue to treat OT vulnerabilities in the same way as IT vulnerabilities, which prevents them from applying risk-based prioritization and focusing on real-world threats. Stronger IT/OT collaboration helps with this prioritization, enables a more comprehensive approach to addressing vulnerabilities, and builds greater confidence when scaling AI.
CEOs must frame industrial AI as a joint operational capability. It is neither an IT project nor an OT experiment. Rather, it is a necessary IT/OT collaboration. As the CISCO survey authors point out, “Organizations confident in scaling AI are those treating infrastructure, cybersecurity, and IT/OT collaboration as foundational, not optional.”
#3. Skills mismatch
For the first time, skills gaps have overtaken headcount shortages as the primary challenge in the industrial cybersecurity workforce, according to a 2026 SANS Institute report. This has real-world impact: 27% of organizations report security incidents directly linked to these capability gaps.
At the same time, AI is creating new opportunities. Nearly half (49 percent) of industrial organizations say AI is reducing manual cybersecurity analysis time and a similar number (48 percent) say AI is automating cybersecurity workflows.
With only 16 percent reporting workforce reduction as a result, AI is creating space to upskill. I believe CEOs should lean into this opportunity to foster an innovation culture. CEOs should embrace a willingness to fail fast, learn from mistakes, question the current status quo and established strategy, and push for greater innovation.
Culture is crucial in AI adoption. Microsoft’s 2026 Work Trend Index shows that organizational factors like manager support and talent practices account for more than 2x the reported AI impact of individual factors like mindset and behavior (67% vs. 32%).
For maximum impact, this culture pivot should be organization wide. AI cannot be the preserve of IT alone. Scaling AI successfully requires a different mindset: CEOs must view AI as a whole organization discipline.
For example, in industry, AI offers a way to accelerate industrial skills training. This is a crucial use case – a report by Deloitte and the Manufacturing Institute estimates that, in the USA alone, mass retirement in the current industrial workforce will create a gap of 3.8 million workers. Writing in Fortune magazine earlier this year, Kriti Sharma argued, “The tech available today is good enough to accurately reflect seasoned experience, shorten learning curves, and close talent gaps with near-instant, but verified and context-rich data gleaned from real-world work… AI can demolish the barrier to entry to industrial jobs without neutering the skills required to do the job.” With a whole-organization approach, these opportunities to leverage AI will be more readily visible and actionable.
#4. Lack of governance
It is said that to build a fast car, you must first give it excellent brakes. But I’d go further than that to say, when done well, AI governance is not a brake on innovation – it is the enabler.
When 60 percent of CEOs admit that they have intentionally slowed AI implementation due to concerns over errors and malfunctions (World Economic Forum), it’s clear good governance must be prioritized earlier.
CEOs must establish responsible AI guidelines and establish AI centers of excellence to promote this knowledge internally. Data governance and compliance is a critical piece of the puzzle. Establishing and enforcing human-in-the-loop checks aids quality and confidence. Compliance implications and vendor selections should be reviewed against relevant standards, including IEC 62443, NERC CIP and NIS2. Zero Trust, role-based access and network segmentation are especially important in OT environments where legacy exposure is high. And, where Agentic AI is used, continual monitoring of the quality and accuracy of performance, results and output should be paired with full lifecycle management; retraining or retiring where appropriate.
As AI handles a greater proportion of operational tasks, human oversight must be maintained with processes to support and enforce it. CEOs must take a lead on this. Deloitte’s State of AI 2026 report found that enterprises where senior leadership actively shapes AI governance achieve significantly greater business value from AI than those delegating the governance work to technical teams alone.
#5. Pilot paralysis
When these foundations are in place, AI can be scaled responsibly and profitably. That’s important because last year an IBM survey found that only 16 percent of AI initiatives scaled enterprise wide.
CEOs must ensure their organizations approach AI not with a pilot mindset but with a programmatic mindset. While pilot success rests on clearly defined objectives, measurable ROI, often with validation before live deployment, and a phased approach, successful scaling demands a strong foundation. And, as we’ve seen here, the vital components of an organization-wide programmatic approach are: data readiness, IT/OT convergence, broad AI skilling, and strong governance – as well as the CEO’s AI vision and leadership.
The next horizon: Agentic AI in industrial operations
2026 is proving to be a turning point in the adoption of AI. It is a defining success measure for CEOs. Twice the number of CEOs (72 percent) identified themselves as their organization’s main decision maker on AI in the BCG AI Radar survey in 2026 than did the year before (37 percent).
But while most see themselves as decision makers, not all CEOs are making the same decisions. BCG categorizes around 15 percent of CEOs as “followers” who recognize AI’s potential but are going slow. Around 70 percent are “pragmatists” – although they take a more active stance, they advance with the market rather than ahead of it. Just 15 percent of CEOs are decisive “trailblazers”. By making AI a top priority, investing at scale, and swiftly upskilling their workforce, trailblazers create a reinforcing cycle: faster adoption, greater confidence, and stronger returns that justify even bolder moves.
Their next move will be the Agentic AI frontier. The biggest opportunities here for industrial firms are identified by Deloitte as supply chain management, predictive maintenance, R&D, cybersecurity, generation dispatch, inventory adjustments, and real-time anomaly responses.
But Microsoft’s 2026 Work Trend Index argues that the rise of AI agents should be understood as more than the next wave of software. It says, “As AI moves from assisting with isolated tasks to participating in workflows across functions and systems, leaders must rethink the fundamental design of the enterprise. Work is no longer organized only around people, processes and applications. Increasingly, it is organized across people, agents and the systems that connect them. The central task of leadership, therefore, is shifting from deploying technology to leading and enabling their teams to redesign work and processes.”
This shift also points to a deeper structural change in how organizations may evolve. AI agents are increasingly likely to be integrated not just as tools but as part of the operational fabric of the enterprise, where they function alongside human teams with clearly defined roles embedded within workflows.
Taken together, this makes one requirement unavoidable: the foundational work must be in place now. Data readiness, IT/OT convergence, workforce skilling, strong governance, and a structured, programmatic approach are becoming prerequisites rather than optional enablers for scaling AI successfully.
