Memo: AI Agents Must Defend Manufacturing Systems Amid Rising Cyberattacks
Integer Uses A Holistic Security by Design Approach for Defense Tech and Software Engineering.
As advanced forms of AI are being used to launch cyberattacks, a recent memo sent to the White House calls for the creation of AI agents (intelligent, autonomous tools) to defend manufacturing systems.
The memo from the Special Competitive Studies Project outlines a strategy to reestablish U.S. leadership in Advanced Manufacturing, citing that by 2030, America’s share of global manufacturing output is projected to decline to just 11 percent. Part of the proposed solution in the memo involves national policies to drive tech adoption such as “establishing testbeds for AI agents to defend manufacturing systems.”
The memo states that:
- Instrumented and realistic evaluations for cyber defense agents that can withstand adversarial attacks are crucial.
- Key actions should include modernizing and scaling testbeds for industrial control systems and IIoT through the National Institute of Standards and Technology (NIST).
- Creating Autonomous Cyber Operations gyms can accelerate AI training and defense against adversarial systems.
I strongly agree with this call to action, though, I believe focusing on AI is just one part of the equation. Engineers and manufacturers must have a holistic security by design approach for end-to-end resiliency in systems, especially as the rate of attacks is faster than ever.
We need to ask ourselves ‘Where is AI important to use and where do we need to step back and fill in our approach with other engineering best practices?’ Use AI as a supplemental tool, not a complete solution.
At Integer, we recognize that AI is a crucial enabler, but it’s equally important to adopt a holistic security approach for defense tech and software engineering that integrates security measures throughout the design and development process.
We are committed to integrating security throughout the entire system lifecycle, from inception to deployment, to build resilient, mission-critical solutions. Our goal is to protect systems from current threats while evolving to address emerging risks, such as agentic AI. By adhering to best practices and standards like post quantum encryption and Zero Trust Architectures, we ensure systems stay ahead of evolving challenges. At the same time, we actively integrate modern security solutions into legacy systems, ensuring their continued viability and security without compromising functionality.
Our advanced manufacturing and cyber-physical resilience capabilities bring together state-of-the-art technologies, blending data, physics, AI, and cutting-edge digital engineering to deliver comprehensive solutions that not only mitigate vulnerabilities but also anticipate and withstand evolving threats.
- Through the use of AI, simulations, and attack emulation, we’re enhancing the resilience of mission-critical unmanned maritime vehicles, robotic systems, and industrial controls for our military and industry customers.
- We empower both systems and operators with the ability to make faster, more informed decisions through enhanced real-time analytics. Our solutions provide you with the data and insights needed to stay ahead of potential security risks.
- By leveraging digital representations of systems, we fuel AI-driven analytics that refine product lines, optimize manufacturing processes, and assess the severity of cyberattacks bringing a holistic understanding of systems vulnerabilities and potential impacts.
As outlined in the memo, we are building Autonomous Cyber Operations gyms — one of the most effective tools for stress-testing systems in a controlled, non-destructive environment. This allows us to simulate real-world cyberattacks and evaluate system vulnerabilities and resilience, without interrupting operations.
It’s not just about setting up testbeds; it’s about ensuring their accuracy. At Integer, our team ensures that our predictive tools are built on the right data and systems representations, using physics based digital engineering, so that decisions made from them are both reliable and actionable.
This rigorous attention to detail ensures that our technologies are engineered with excellence, enabling faster, more informed decisions and better long-term outcomes.