Report urges AI critical infrastructure designation under CISA
The AI sector's concentration and interdependence already mirror critical infrastructure, but the recommendation lands in an administration that has explicitly rejected mandatory frameworks for the industry.
TL;DR
A report from Americans for Responsible Innovation, shared exclusively with CyberScoop, calls on the federal government to designate the AI sector as critical infrastructure and name CISA as its lead risk management agency. The report defines the sector broadly to include frontier model weights, datacenters, AI-specific hardware, and semiconductor chips, and argues a single attack on the AI stack could cascade across multiple sectors. That argument isn't new, but it lands in an administration that's explicitly rejected mandatory frameworks for the AI industry.
The report, authored by Terrence Kelly and Jessica Maksimov at Americans for Responsible Innovation, defines the AI sector to include frontier model weights, evaluation and alignment systems, datacenters, AI-specific hardware, semiconductor chips, and the platforms used to deploy models at scale. It argues CISA is the natural lead agency: the agency already manages eight critical infrastructure sectors and has the statutory coordination authority to work across the interagency on threats that don't respect sector boundaries.
The threat picture lends weight to the argument. Iranian drones struck Amazon-owned datacenters last year. AWS, Microsoft Azure, and Google Cloud together control 63% of the U.S. datacenter market, creating a concentration risk where a single disruption could cascade across multiple sectors at once. The UK has already designated datacenters as a standalone critical infrastructure sector.
But the recommendation lands in an administration that's been unambiguous about its aversion to mandatory frameworks. EO 14409, signed in June, explicitly states that nothing in the government's AI testing program 'will be construed as mandatory or part of a federal licensing or permitting regime.' AI czar David Sacks called the scaled-back order, which reduced government pre-release access to models from 90 days to 'up to' 30, 'a game changer.' The administration's alternative approach runs through voluntary channels: the Gold Eagle vulnerability clearinghouse, secure-by-design principles, and an AI-ISAC for threat-intelligence sharing. None of it carries the force of a critical infrastructure designation, which unlocks federal prioritization and mandatory security requirements the White House hasn't shown any interest in imposing.
Published ·Deep Fathom