By Steven Sim Kok Leong, Co-Founder, AI GRC SG Community
The convergence of Operational Technology (OT), Artificial Intelligence (AI), and emerging Quantum Computing technologies represents one of the most significant technological shifts facing critical infrastructure operators today. Energy networks, logistics hubs, transportation systems, ports, manufacturing facilities, and industrial environments increasingly rely on interconnected OT systems that were never designed to operate alongside autonomous AI agents, machine learning models, or future quantum-capable technologies. While this convergence promises unprecedented operational efficiency, predictive capabilities, and autonomous decision-making, it also introduces new categories of cyber, safety, governance, and operational risks that traditional technology governance frameworks are not equipped to manage.
The Convergence Challenge
OT environments were historically designed around principles of stability, deterministic behavior, long asset lifecycles, and safety. Many industrial control systems, programmable logic controllers (PLCs), supervisory control and data acquisition (SCADA) systems, and distributed control systems (DCS) remain operational for decades. Their primary objective is reliability rather than adaptability.
In contrast, AI systems are dynamic, probabilistic, and constantly learning. AI introduces autonomous decision-making, predictive analytics, generative capabilities, and machine-speed actions into environments where predictability is critical. Quantum computing, while still emerging, introduces another level of disruption by potentially rendering current cryptographic protections obsolete while enabling unprecedented optimization and analytical capabilities.
The intersection of these technologies creates a fundamental mismatch. Legacy OT systems expect stable and predictable interactions, whereas AI and quantum technologies thrive on continuous adaptation and complex computation. OT-ISAC captures the operational challenge succinctly: “The gap isn’t intelligence. It’s the action.”
The mismatch must therefore be governed not only as a technology risk, but as a decision-making problem at the cyber-physical boundary.
Key Risk Areas
1. Safety Risks from AI-Driven Decisions
In traditional IT environments, an erroneous AI recommendation may create an operational inconvenience. In OT environments, a faulty AI decision could result in equipment damage, environmental incidents, physical injuries, or service disruptions.
AI models embedded into industrial operations may make decisions based on incomplete, inaccurate, or manipulated data. Model hallucinations, inaccurate predictions, or inappropriate autonomous actions can directly affect physical processes. As organizations move toward autonomous operations, the distinction between cyber incidents and safety incidents becomes increasingly blurred. This reinforces ISACA research that responsible AI requires policy, impact assessment, security and privacy safeguards, and lifecycle management rather than isolated technical controls.
A predictive maintenance AI, for example, might incorrectly determine that a critical asset is healthy, delaying maintenance and potentially leading to catastrophic failure.
2. Expanded Cyber Attack Surface
The introduction of AI significantly increases the attack surface of OT environments. Attackers may target not only traditional OT assets but also:
· AI models
· Training datasets
· Inference engines
· AI agents
· Data pipelines
· Model orchestration platforms
· AI gateways
Adversaries can conduct model poisoning, prompt injection, adversarial machine learning attacks, and data manipulation campaigns. In OT environments, these attacks can influence operational decisions rather than merely compromise information systems.
The risk is further amplified because many legacy OT assets lack strong authentication, encryption, and monitoring capabilities. OT-ISAC research highlights the need to convert community-enriched signals into operational decisions, asking whether a threat affects the operator, requires action and could alter operational impact.
3. Quantum Threats to Industrial Security
Quantum computing presents both opportunities and threats.
Most current industrial systems rely on public-key cryptographic algorithms that may become vulnerable to sufficiently powerful quantum computers. While large-scale cryptographically relevant quantum computers remain under development, industrial assets often have lifecycles exceeding twenty years, meaning systems deployed today may still be operational when quantum threats become practical.
Organizations face a "harvest now, decrypt later" risk, where adversaries capture encrypted communications today for future quantum-enabled decryption. OT operators managing critical infrastructure, intellectual property, safety data, and operational commands must therefore begin planning migration to quantum-resistant cryptography.
4. Autonomous Agent Risks
The emergence of agentic AI creates a unique governance challenge. Unlike traditional analytical AI, AI agents can plan, reason, execute tasks, and interact with multiple systems autonomously.
When connected to OT environments, agents may:
· Change system configurations
· Trigger workflows
· Generate operational instructions
· Interact with industrial control platforms
· Coordinate multiple automated actions
Without appropriate constraints, a single agent failure could propagate rapidly across interconnected systems, creating systemic operational disruptions. ISACA’s 2026 research warns that organisations are “deploying AI faster than they can govern it”. For OT, this makes tested intervention, emergency-stop and recovery mechanisms essential before agents receive operational authority.
5. Supply Chain and Third-Party Risks
Modern AI capabilities often rely on external models, cloud providers, APIs, and specialist vendors. OT operators therefore become dependent on a broader ecosystem of suppliers.
This introduces risks related to:
· Model provenance
· Vendor security practices
· AI service availability
· Software supply chain compromise
· Hidden dependencies
A compromised AI vendor could indirectly influence industrial decision-making processes across multiple organizations simultaneously.
Governance Implications
The convergence of OT, AI, and quantum technologies demands a significant evolution in governance structures.
Traditional IT governance often focuses on confidentiality, integrity, and availability. OT governance additionally prioritizes safety, reliability, resilience, and operational continuity. AI governance introduces fairness, accountability, transparency, explainability, and human oversight. Quantum governance introduces cryptographic resilience and long-term technology transition planning.
Organizations must therefore establish integrated governance that addresses all four domains simultaneously. This aligns with ISACA’s Digital Trust Ecosystem Framework, which applies systems thinking across security, privacy, quality, resilience, ethics and governance.
Key governance principles should include:
· Safety-by-design
· Security-by-design
· Human-in-the-loop controls
· Explainable AI requirements
· Quantum readiness planning
· Lifecycle governance
· Continuous risk assessment
Board-level oversight is increasingly necessary because failures in these environments may generate enterprise-wide consequences affecting operations, reputation, safety, and regulatory compliance. ISACA’s 2026 AI Pulse Poll found that only 38 percent of surveyed organisations had a formal, comprehensive AI policy, while 56 percent of respondents did not know how quickly their organisation could halt an AI system after a security incident.
Governance Operations (GovOps) for Converged Environments
Traditional governance models are often too slow for rapidly evolving AI and cyber threats. Organizations require Governance Operations (GovOps) capabilities that continuously monitor, assess, and enforce governance requirements.
A modern GovOps function should perform several critical responsibilities.
First, establish a centralized inventory of OT assets, AI models, AI agents, and quantum-sensitive systems. Organizations cannot govern what they cannot identify.
Second, conduct continuous risk assessments rather than annual reviews. AI models, datasets, and threat landscapes evolve constantly. OT-ISAC’s Predictive Resilience work describes the need to “recognize weak signals, challenge assumptions and make informed decisions” before the full incident picture is available. This requires joint interpretation of security monitoring, engineering systems, vendor activity, operational observations and physical processes.
Third, enforce approval gates for AI deployment into safety-critical environments. Not every AI use case should be permitted in OT operations.
Fourth, monitor model performance, drift, explainability, and safety outcomes throughout the AI lifecycle. Consistent with ISACA research, monitoring should also preserve evidence for accountability, investigation, audit and explanation to leadership or regulators.
Fifth, establish quantum transition roadmaps to identify cryptographic dependencies and prioritize migration toward post-quantum cryptography.
Finally, create integrated assurance processes combining cybersecurity, engineering, safety, risk management, compliance, and AI governance expertise. In practice, this creates a Digital Trust Operations capability that combines OT-ISAC’s operational context and collective resilience with ISACA’s lifecycle governance, accountability and assurance principles.
The combined research points to five connected operational capabilities: OT resilience operations, AI governance operations, cybersecurity operations, quantum-readiness operations, and digital-trust assurance. These should operate as one continuous governance system rather than as separate oversight functions.
The Path Forward
The future industrial enterprise will increasingly operate at the intersection of OT, AI, and quantum technologies. Organizations that view these domains separately risk creating governance blind spots and operational vulnerabilities. The challenge is not merely technological but organizational. Success requires integrated governance that balances innovation with safety, autonomy with accountability, and digital transformation with operational resilience.
The organizations that thrive in this environment will be those that establish effective governance operations today, ensuring that emerging technologies enhance industrial performance without compromising the safety, reliability, and trust upon which critical infrastructure ultimately depends. OT-ISAC’s emphasis on evidence-based operational decisions under uncertainty, together with ISACA’s focus on digital trust and lifecycle governance, provides a practical foundation for this integrated model.
References
[1] OT-ISAC, “Operational Resilience” and mission statement, https://www.otisac.org/
[2] OT-ISAC, NCSA Thailand and TXOne Networks, “Predictive Resilience Exercise,” reported 14 September 2026,
[3] OT-ISAC, “Resources,” including OT threat reports, vulnerability advisories and Bridging the Gaps in OT Cybersecurity Programs, https://www.otisac.org/resources
[4] ISACA Journal, “How to Build Digital Trust in AI With a Robust AI Governance Framework,” 1 May 2024,
https://www.isaca.org/resources/isaca-journal/issues/2024/volume-3/how-to-build-digital-trust-in-ai-with-a-robust-ai-governance-framework
[5] ISACA, “Digital Trust Ecosystem Framework,” https://www.isaca.org/digital-trust
[6] ISACA, “AI Use Accelerates, While Governance and ROI Lag, Says New ISACA Research,” 5 May 2026,
https://www.isaca.org/about-us/newsroom/press-releases/2026/ai-use-accelerates-while-governance-and-roi-lag-says-new-isaca-research
[7] ISACA, “New ISACA Research: 56% of Digital Trust Pros Don’t Know How Fast They Could Shut Down AI After a Security Incident,” 23 March 2026, https://www.isaca.org/about-us/newsroom/press-releases/2026/digital-trust-pros-dont-know-how-fast-they-could-shut-down-ai-after-a-security-incident
[8] ISACA, “New ISACA Research: Quantum-Resistant Cryptography Not a Matter of ‘If’ but ‘Right Now’,” 13 October 2026, https://www.isaca.org/resources/news-and-trends/isaca-now-blog/2023/quantum-resistant-cryptography
Latest News
