The "Roles and Responsibilities Framework for Artificial Intelligence in Critical Infrastructure," developed with DHS Secretary Alejandro N. Mayorkas and the new AI Safety and Security Board, aims to guide AI integrators — including cloud providers, developers and critical infrastructure operators — in collaborating on safe AI practices across essential sectors like energy, water and telecommunications.
Specifically, the written framework will address three aspects identified by DHS as core risk areas facing critical infrastructure in today’s society: AI-driven attacks, targeted attacks on AI systems and design flaws. The framework distinguishes the unique challenges each AI stakeholder may encounter and speaks to specific recommendations for each group.
For cloud and compute providers, it emphasizes securing AI development environments, including access management and physical data center security, while encouraging monitoring for anomalies and pathways to report suspicious activities.
AI developers are urged to follow a secure-by-design approach, ensuring models align with human values, evaluating biases and failure risks, and supporting independent assessments.
Critical infrastructure operators are advised to implement strong cybersecurity for AI deployments, protect customer data, maintain transparency in AI use and actively monitor AI system performance to inform developers about potential impacts.
Civil society groups, universities and consumer advocates are encouraged to participate in standards development, research AI’s societal impact, and help shape AI’s safe deployment in critical infrastructure. Meanwhile, public-sector entities on local and federal levels are tasked with promoting responsible AI use, supporting standards development and collaborating internationally to protect those that might be impacted globally, according to the AI guidelines.
Mayorkas underscored the transformative potential of AI for national infrastructure while emphasizing a delicate, balanced approach to the technology.
“AI offers a once-in-a-generation opportunity to improve the strength and resilience of U.S. critical infrastructure, and we must seize it while minimizing its potential harms. The framework, if widely adopted, will go a long way to better ensure the safety and security of critical services that deliver clean water, consistent power, Internet access and more,” Mayorkas said in a DHS news release discussing each sector’s joint goal. “The choices organizations and individuals involved in creating AI make today will determine the impact this technology will have in our critical infrastructure tomorrow.”
At a Thursday press briefing, Mayorkas shed more light on the makeup of the AI Safety and Security Board, formed earlier this year, and how their overall objectives for AI implementation across the nation are taking shape.
“In forming the board, I sought members who are leaders in their fields and who collectively would represent each integral part of the ecosystem that defines AI's deployment in critical infrastructure,” he said. “We have … assembled such a board comprised of leaders of cloud and compute infrastructure providers, AI model developers, critical infrastructure owners and operators, civil society and the public sector. We believe the safety and security of our critical infrastructure is a shared responsibility.”
The framework emerges as AI increasingly contributes to resilience and risk mitigation across sectors. It is already enhancing earthquake detection, stabilizing power grids and sorting mail, according to the DHS news release.
The department also must take international implications into account, as the technology expands well beyond U.S. borders. DHS is seeking to harmonize AI standards internationally.
“We have spoken about the fact that we not only want to ensure that these guidelines are adopted and implemented domestically but also across the Atlantic, internationally,” Mayorkas added, noting with the EU advancing its own AI regulations, the DHS framework could play a pivotal role in aligning efforts worldwide.
However, the primary focus for the U.S. currently is ensuring the newly released AI guidelines are adopted across the nation. Although the framework is voluntary, he said DHS hopes a show of support across agencies will encourage rapid adoption.
“We expect the board members to implement the guidelines, to catalyze other organizations in their respective spheres and across the ecosystem, to adopt and implement the guidelines as well,” Mayorkas said. Should it be adopted broadly, he said the framework may help to delay or prevent premature regulations that could “impair our leadership in the world and suffocate our inventiveness.”
