STAISYS uses emulated energy environments to evaluate AI systems against realistic cyber threats at scale
Artificial intelligence has the potential to strengthen the security, resilience and reliability of America’s energy sector. As utilities, grid operators and technology providers explore AI-enabled tools for energy operations and cybersecurity, these systems must be tested in realistic environments before they are integrated into critical infrastructure.
The U.S. Department of Energy’s Office of Cybersecurity, Energy Security and Emergency Response, in partnership with Sandia National Laboratories, is advancing Scaled Testing for AI Systems, or STAISYS, a project designed to evaluate AI technologies in emulated cyber-physical environments relevant to critical energy infrastructure. STAISYS supports CESER’s Artificial Intelligence For Operationally Resilient Technologies and Systems (AI-FORTS) program by helping determine how AI systems perform when exposed to realistic operating conditions and cyber threats. STAISYS is one of four Sandia-led projects funded through AI-FORTS.
AI technologies developed for energy security must be evaluated not only as individual components, but as part of the larger systems into which they are integrated because adversaries may target any part of that environment. These systems can include operational technology, industrial control systems, communications networks, sensors, analytics platforms and other cyber-physical infrastructure.
STAISYS addresses this challenge through emulytics, an approach that combines emulation and analytics to create realistic, repeatable test environments. These environments allow researchers to integrate AI technologies into representative energy systems, emulate cyber activity and run experiments at scale. The project is designed to help CESER and its partners better understand AI system maturity, security and readiness for potential operational use.
The STAISYS approach brings together three core capabilities:
- Integration of AI technologies into emulated testbeds that represent critical infrastructure systems and their supporting cyber-physical environments.
- Development of realistic threat models and AI kill chains to examine how adversaries could target AI-enabled technologies, supporting infrastructure or the interactions between them.
- Statistical evaluation of experiments at scale to quantify potential impacts, characterize uncertainty and compare how different attack vectors affect both AI performance and energy system operations.
This scaled testing approach helps move beyond one-time demonstrations or isolated model evaluations. By running many experiments across varying configurations and threat conditions, STAISYS can provide deeper insight into when an AI capability performs as expected, where it may be vulnerable and how those vulnerabilities could affect the broader system.
A proof-of-concept demonstration has already been developed to evaluate the security of an AI-enabled cyber-physical intrusion detection capability for an industrial control system. Building on this foundation, STAISYS will support the development of emulated environments aligned with CESER’s near- and long-term AI-FORTS priorities. The project will also support collaboration across the national laboratories to incorporate emerging AI technologies and identify representative testbed environments.
The results will help inform decisions about AI technology readiness, security risk and future research and development investments. For utilities, grid operators, technology developers, national laboratories and other qualified stakeholders, STAISYS offers a pathway to better understand how AI-enabled energy security technologies behave under realistic and repeatable conditions.
Cybersecurity of critical energy infrastructure is central to the development of STAISYS. By evaluating AI systems in the context of their intended cyber-physical environments, CESER and Sandia are helping ensure that AI technologies for the energy sector are not only innovative, but secure, resilient and ready to address evolving threats to the nation’s critical infrastructure and national security.
Learn more about Sandia’s Electric Grid Cyber and Physical Security work.
For more information about STAISYS, contact Deepu Jose, djose@sandia.gov, or Meg Sahakian, mgaliar@sandia.gov.
October 19, 2026