Scaling industrial AI is more a human than a technical challenge

Industrial artificial intelligence has moved from promise to practice. A Cisco Systems Inc. survey of 1,000 industry leaders found that 61% of organizations across manufacturing, transportation and utilities are deploying AI to improve productivity, reduce costs and strengthen operational resilience.

Yet despite this momentum, only 20% have reached truly scaled, mature adoption. The technology is advancing rapidly, but progress often slows once AI moves beyond pilots and into production. Lack of necessary infrastructure, cybersecurity risk and system complexity are the most common reasons, but beneath these technical challenges lies a more fundamental constraint, one rooted in how people work together.

Industrial AI is a team sport

Industrial AI sits at the intersection of two disciplines with very different histories. Information technology teams are trained to manage networks, data, security and digital platforms. Operational technology teams are experts in industrial processes, safety, reliability and real‑time operations. Both bring essential capabilities, but neither can scale AI alone.

AI doesn’t replace this division of expertise but amplifies it. As AI systems connect more assets, move decisions closer to operations and increase reliance on data, the need for coordination grows. When IT and OT operate in silos, organizations struggle to deploy AI confidently in production, regardless of how advanced the technology may be.

Cisco’s survey found that while 57% of organizations report some level of IT/OT collaboration, a significant 43% minority operate with limited or no meaningful cooperation. Fully converged teams remain rare. This isn’t because leaders don’t recognize its value but because building combined IT/OT skill sets in individuals is difficult and often unrealistic.

Collaboration, not convergence

Expecting individuals to master both IT and OT disciplines is rarely practical. The combined skill set is unusual. What matters far more is enabling collaboration, creating environments where IT and OT teams can bring their full expertise to the table and work toward shared outcomes.

Organizations that enable this kind of collaboration report higher confidence in their ability to scale AI. They also experience greater network stability and place stronger emphasis on cybersecurity as a foundational requirement, rather than an afterthought. In contrast, organizations with segregated teams are more likely to experience instability, slower deployment and elevated risk.

This is a human challenge as much as a technical one. It requires trust, shared language and aligned incentives. It also requires leadership to frame AI not as an IT project or an OT experiment, but as a joint operational capability.

Addressing risk

As AI expands connectivity and data flows, cybersecurity concerns rise sharply: Forty percent of organizations cite it as the single biggest obstacle to scaling industrial AI, and 48% identify security and segmentation as their top networking challenge. Organizations with stronger IT/OT collaboration are more likely to recognize these risks early and address them collectively.

Where silos persist, risk is often fragmented. OT teams may prioritize availability and safety, while IT teams focus on security controls and compliance. Without collaboration, tradeoffs are harder to manage, and AI deployments remain limited to lower‑risk environments. By working together, teams can design systems that balance security with operational continuity, a prerequisite for deploying AI in production environments where failure is not an option.

Accountability builds confidence

Organizations that struggle to scale AI often hesitate not because the technology is unproven, but because ownership is unclear. Who is accountable when an AI‑driven system affects operations? Who responds when performance degrades or security alerts appear?

Organizations further along in their AI journey tend to address these questions through shared governance and clearer accountability across IT and OT. This doesn’t require a structural overhaul, but rather agreement on common goals: uptime, safety, resilience and performance.

Over time, this collaboration also supports workforce readiness. Skills shortages remain a barrier, cited by 34% of organizations overall, but this drops to 27% among more mature AI adopters, suggesting that experience and collaboration help close the skills gap over time.

Ultimately, realizing the full potential of Industrial AI requires dismantling silos so that IT and OT teams can bring their distinct competencies to a shared table. The goal is not to engineer a rare breed of hybrid super-worker, but to forge truly connected teams. By seamlessly combining digital agility with operational rigor, these unified teams turn AI’s promise into sustained, everyday impact.

Image: SiliconANGLE/Ideogram

Samuel Pasquier is vice president of product management for industrial IoT networking at Cisco Systems. He wrote this article for SiliconANGLE.