Rockwell Automation has announced its participation in Anthropic’s Project Glasswing, an initiative exploring the use of advanced AI for defensive cybersecurity across critical software and infrastructure. Through controlled access to Claude Mythos 5, Rockwell security teams are evaluating how AI-assisted analysis can accelerate the discovery, validation, prioritization and remediation of vulnerabilities across software and connected systems used within modern manufacturing environments.

The collaboration comes as industrial operations become increasingly connected across production equipment, control systems, software platforms, operational data and enterprise infrastructure. That connectivity creates significant opportunities for greater visibility, automation and production efficiency, but it also expands the number of systems, interfaces and software dependencies that must be secured. For manufacturers, cybersecurity is therefore no longer separate from operational performance. It is increasingly tied to equipment availability, production continuity, safety and overall plant resilience.

Robotic hand using touchscreen representing AI-enabled industrial cybersecurity | Industrial Automation Monthly

Rockwell Automation is participating in Anthropic’s Project Glasswing to explore AI-assisted vulnerability discovery and remediation across software and connected industrial systems. (Components Source Network editorial stock photo)

“Rockwell Automation plays a central role in helping manufacturers operate safely, securely, and reliably,” said Tony Baker, vice president and Chief Product Security Officer, Digital Trust, Rockwell Automation. “By participating in Project Glasswing, we are applying advanced AI capabilities in a controlled, defensive way to help identify and address vulnerabilities faster and support the resilience of the industrial ecosystem.”

Anthropic launched Project Glasswing in April 2026 with approximately 50 initial partners using Claude Mythos Preview for defensive cybersecurity work. The initiative was subsequently expanded to approximately 150 additional organizations across more than 15 countries, including organizations supporting critical infrastructure in power, water, healthcare, communications and hardware. Rockwell is participating through controlled access to Claude Mythos 5, applying the technology specifically within a defensive industrial cybersecurity context.

That industrial context matters. Vulnerability management in operational technology (OT) environments can be considerably more complex than in conventional enterprise IT. Industrial control systems may support continuous production, safety-critical operations and assets expected to remain in service for years or even decades. Identifying a vulnerability does not necessarily mean a patch can be deployed immediately. Manufacturers may first need to assess its operational relevance, validate the proposed remediation, test compatibility with existing systems, coordinate maintenance windows and consider the potential impact of any change on production or safety.

For that reason, faster vulnerability discovery represents only one part of the cybersecurity challenge. Security and engineering teams must still determine which findings create meaningful operational risk, distinguish actionable issues from lower-priority vulnerabilities, and decide how remediation can be implemented without introducing new instability into functioning production systems. In some cases, patching may be appropriate. In others, segmentation, access controls, monitoring or other compensating measures may be required until a permanent remediation can be introduced safely.

This is where AI-assisted defensive analysis could provide practical value. The ability to examine software, identify potential weaknesses and help security teams validate and prioritize findings can reduce some of the manual effort associated with the early stages of vulnerability management. Within industrial environments, however, those capabilities still need to operate alongside established cybersecurity processes, engineering review, change management and an understanding of the physical processes being controlled.

Rockwell’s participation in Project Glasswing therefore represents more than the application of AI to vulnerability scanning. It reflects a broader shift toward using advanced analytical tools within industrial cybersecurity while preserving the engineering discipline required by operational environments. As manufacturing systems become more software-defined and interconnected, cybersecurity teams will increasingly need tools capable of keeping pace with that complexity without losing sight of the operational consequences behind each security decision.

For manufacturers, that balance is especially important. Cyber resilience ultimately depends not only on finding vulnerabilities faster, but on understanding which vulnerabilities matter most, how they affect production environments and how they can be addressed without compromising the availability, reliability and safety those systems were designed to maintain.

About Rockwell Automation

Rockwell Automation develops industrial automation and digital transformation technologies used across manufacturing and production environments. Headquartered in Milwaukee, Wisconsin, the company serves customers in more than 100 countries and employed approximately 26,000 people as of the end of fiscal year 2025. For more information, please click here. 

About Anthropic

Anthropic develops artificial intelligence systems and the Claude family of AI models. Its Project Glasswing initiative provides approved organizations with controlled access to advanced AI capabilities for defensive cybersecurity research and vulnerability discovery. To learn more, please click here. 

Source: Rockwell Automation

(Editor’s Note: All trademarks mentioned in this article, including company names, product names, and logos, are the property of their respective owners. Use of these trademarks is for informational purposes only and does not imply any endorsement.)

Molly Bakewell Chamberlin
Latest posts by Molly Bakewell Chamberlin (see all)
Tagged