A recent warning from the Department of Homeland Security about potential Iranian cyber activity highlighted long-standing vulnerabilities in U.S. energy systems. Experts note these weaknesses predate current artificial intelligence concerns and stem from infrastructure built before modern digital threats emerged.

Joshua Corman, executive in residence for public safety and resilience at the Institute for Security and Technology, stated that critical energy systems were "always prey" and had been "surviving at the appetite of our predators" for years. Speaking about recent incidents involving AI tools, Corman emphasized that while AI increases offensive capabilities, human adversaries remain the central concern.

"Any sociopath that wants to [attack] is now more powerful than they used to be," Corman said. "This has been a force multiplier and continues to grow."

Cybersecurity professionals caution against overstating AI's role in imminent threats. Rob Denaburg, senior manager for cybersecurity at the American Public Power Association, noted that documented cases of AI-driven attacks on energy infrastructure remain limited. In a recent incident involving OpenAI's model breaching Hugging Face, attackers focused on fulfilling predefined objectives rather than targeting critical systems directly.

Nation-state actors have historically posed the greatest risk to power grid security, according to Corman. However, AI tools now enable less-skilled adversaries to execute more sophisticated attacks by automating tasks like identifying and exploiting network weaknesses. Corman explained that generative AI can "read the manuals" for industrial control systems, allowing attackers without specialized knowledge to launch effective campaigns.

Utilities are adopting defensive strategies such as manual override capabilities and network segmentation to mitigate risks. Sophie McDowall, research associate at the Foundation for Defense of Democracies, noted that AI development is "partially causing" the vulnerabilities it amplifies. OpenAI's recent $1 billion commitment to AI security research was cited as a positive step, though experts stress regulatory frameworks for AI in critical infrastructure remain underdeveloped.

Corman warned against deploying AI solutions too rapidly in operational technology environments, describing the scenario as "an AI bull fighting another AI bull in an OT china shop." The core challenge, experts agree, is maintaining system resilience regardless of whether threats originate from human or artificial sources.