Honeywell’s 2026 OT Cybersecurity Benchmark Report, released on Tuesday, reveals significant disparities between perceived and actual preparedness of industrial organizations’ operational technology (OT) security frameworks. The report also delves into the influence of artificial intelligence (AI) on OT security.
Current State of OT Security
In a survey involving 603 leaders from critical infrastructure sectors, 88% rated their OT security programs as either mature or design-led. However, only 21% reported maintaining a comprehensive inventory of their OT assets, underlining a gap between perception and reality.
The respondents were drawn from diverse sectors, such as energy, oil and gas, healthcare, maritime, and manufacturing, with participants from the Americas, EMEA, and APAC regions.
Challenges in Monitoring and Incident Management
Visibility issues are apparent in both asset inventory and monitoring. Only 33% of those surveyed indicated full integration of OT systems into a centralized security operations center. Furthermore, a mere 20% continuously monitor over three-quarters of their IoT devices.
Organizations that faced significant OT cybersecurity incidents reported an average downtime of 16.2 hours. Among these, 21% estimated their downtime costs exceeded $100,000 per hour, while 4% reported losses surpassing $500,000 per hour.
Incident occurrence varied across sectors, with 91% of energy and utilities and 87% of maritime respondents reporting significant incidents in the past year, compared to 54% in the oil and gas sector. In the healthcare sector, only 19% confirmed that their facility systems are fully integrated into cybersecurity monitoring.
The Role of AI in OT Security
AI is expected to significantly impact OT security operations within the next 2-3 years, according to 99% of respondents. Currently, AI-enabled tools are widespread, with 72% employing AI for threat detection, 68% for continuous monitoring, and 59% for asset inventory management.
Despite these advancements, just 23% are using autonomous AI for threat detection, indicating that most current AI applications support human analysts rather than function independently.
Honeywell’s report emphasizes the need for clear decision-making frameworks and human oversight as AI evolves from assisting to taking autonomous actions. The goal is to enhance visibility and response capabilities without introducing new operational risks.
This report highlights the ongoing challenges and opportunities in integrating AI into OT security. As organizations aim for greater efficiency and reliability, the effective governance of AI tools will be crucial in mitigating potential risks.
