- Agentic AI now represents top attack vector for 48% of cybersecurity professionals
- Manufacturing accounted for 27.7% of cybersecurity incidents in 2025
- Cyber ranges provide safe environments for testing AI-powered security responses
- Production disruptions from cyberattacks extend beyond data theft to operational impact
As manufacturing plants increasingly deploy autonomous AI agents for predictive maintenance and process optimization, nearly half of cybersecurity professionals now believe agentic AI will represent the top attack vector for cybercriminals and nation-state threats by the end of 2026. This shift has prompted manufacturers to adopt cyber range training environments where security teams can practice defending against attacks that target both information technology systems and operational technology on the factory floor. The manufacturing industry accounted for 27.7% of cybersecurity incidents in 2025, according to IBM’s 2026 X-Force Threat Intelligence Index, making defensive preparation more critical than ever.
Why Are AI Agents Creating New Attack Surfaces?
Every AI agent introduced into an environment creates new access points, new authentication challenges, and new pathways for attackers to exploit. Unlike traditional software that follows predetermined logic, agentic AI systems adapt their behavior based on context, learn from interactions, and increasingly operate with delegated authority across enterprise systems. The manufacturing sector faces particular vulnerability because they’re adopting it to streamline operations, to implement things like predictive maintenance and smart manufacturing.
A manufacturing company’s procurement agent was manipulated over three weeks through seemingly helpful “clarifications” about purchase authorization limits, demonstrating how persistent prompt injection attacks can compromise autonomous systems. When an agent can act autonomously, mistakes propagate faster, blast radius increases, and rollback becomes harder. The integration of industrial control systems with AI agents means a single compromised system can disrupt production lines, not just compromise data.
How Do Cyber Ranges Address Manufacturing Security Challenges?
A cyber range is a virtual environment used for computer security, cyberwarfare training, simulation or emulation, and development of technologies related to cybersecurity. For manufacturing environments, these platforms replicate both IT (information technology) and OT (operational technology) systems, allowing security teams to practice responses without risking production equipment. The NCRC staff applies its cybersecurity and cyber engineering expertise to provide event environments with the necessary fidelity and realism to support cyber-related testing and the development of cyber-related tactics, techniques, and procedures.
Modern cyber ranges enable manufacturers to test how people, tools, processes, and AI-powered security agents perform under simulated attack conditions. They’re moving past the office network to disrupt the operational technology (OT) systems that keep your machines running, making it essential for security teams to understand both IT and OT vulnerabilities. More than 22% of organizations reported a cybersecurity incident affecting OT systems in the past year, with 40% of these incidents causing operational disruption.
What Security Controls Work for Autonomous AI Agents?
Organisations should assume that agentic AI systems may behave unexpectedly and plan deployments accordingly, prioritising resilience, reversibility and risk containment over efficiency gains. Western governments have issued guidance emphasizing that strong governance, explicit accountability, rigorous monitoring and human oversight are essential prerequisites for deploying autonomous agents in manufacturing environments.
Securing these agents requires the same rigor and traceability applied to human users, yet organizations are deploying hundreds of AI agents, yet they lack agentic identity governance policies to help manage them safely. Security for agentic AI relies on defense in depth, requiring manufacturers to implement controls at multiple levels: model security, application architecture, identity and access management, and infrastructure policy enforcement. Leadership focus is shifting toward cyber resilience: Maintaining production continuity under attack.
Manufacturing leaders must treat agentic cybersecurity as an operational continuity issue, not just an IT concern. Investing in cyber range training enables your security teams to practice defending against AI-powered attacks in realistic scenarios before they impact production. Prioritize identity governance for AI agents with the same rigor you apply to human users, implement defense-in-depth strategies across IT and OT systems, and plan deployments with resilience and reversibility built in from day one.
Q: How do agentic AI systems differ from traditional automation in terms of security risk?
Traditional automation follows predetermined logic paths, while agentic AI systems make autonomous decisions, adapt behavior based on context, and can execute multi-step tasks across enterprise systems. This autonomy means compromised agents can chain actions across platforms at machine speed, creating privilege escalation risks that don’t exist with static automation.
Q: What should plant managers look for when evaluating cyber range training for manufacturing environments?
Ensure the cyber range can replicate both IT and operational technology systems with realistic fidelity, including SCADA (Supervisory Control and Data Acquisition), PLCs (Programmable Logic Controllers), and industrial control systems. Look for platforms that support testing scenarios where attacks impact production equipment, not just data systems, and that allow teams to practice coordinated responses involving both IT security and operations personnel.
Article Source: The State of Agentic Cybersecurity
