OpenAI is spending $1 billion to help critical infrastructure defend itself with AI. That money won't arrive as direct payments — it's being converted into Daybreak model access, training, and technical support.
The first batch is aimed at U.S. water utilities, electric grids, local governments, community banks, nonprofits, and open-source projects, with the money allocated over the next six months.
Daybreak helps those organizations find vulnerabilities, analyze anomalies, validate risks, and test patches. OpenAI says roughly 2,000 organizations and workspaces are already using it, and the company plans to plug it into more than 35 enterprise products and services.
The program lands on the same day as GPT-6 Astra, OpenAI's first model to hit the Critical threshold for cybersecurity capability. With the right tools, Astra can already hunt for unknown vulnerabilities and investigate new ways to exploit them without step-by-step human guidance. OpenAI expects AI-driven attacks to become more common and more sophisticated in the coming months.