OpenAI committed $1 billion in subsidized access to its Daybreak AI cyber platform on September 3, 2026, targeting resource-constrained operators of water and wastewater systems, electric grids, state and local governments, community banks, and nonprofits. The commitment is targeted to be consumed over six months and is one thousand times larger than a prior $1 million offer OpenAI made after attacks on U.S. water systems. The initiative accompanied an open letter signed by more than 100 companies, including Anthropic, AWS, Google, Microsoft, and Oracle, calling on governments to fund and expand AI-enabled cyber defense for critical infrastructure. The letter warns of a narrowing window for defenders to patch vulnerabilities before a tsunami of AI-enabled attacks exploits them first. SpaceX and Nvidia did not sign; AWS signed though Amazon did not fully.
OpenAI is running a pilot with the Multi-State Information Sharing and Analysis Center (MS-ISAC) to train state, local, tribal, and territorial cyber defenders on Daybreak tools. One analyst said the MS-ISAC training component matters more than the $1 billion headline, warning that without trained personnel, organizations risk acting on AI recommendations they cannot properly evaluate. Daybreak is described as an agentic defense loop covering inventory, discovery, validation, ownership assignment, remediation, and proof of fix, integrating Codex Security, GPT Cyber models, and the Daybreak Defense Network. Reports on the network's size differ: one source put it at more than 350 enterprise products and partner-operated services, while OpenAI's own announcement described more than 35 joining the network. OpenAI also published its Defense Factory architecture for other defenders to adapt, and simultaneously reversed its prior opposition to California's SB 53, calling for the bill to be strengthened.