OpenAI launches $1B initiative to arm cyber defenders protecting critical infrastructure

OpenAI president Greg Brockman unveiled Daybreak for Frontline Defenders this week, a new global program designed to give cybersecurity professionals access to frontier AI models and defensive tooling to safeguard critical services. The announcement represents a significant expansion of the company's existing Daybreak cyber defense platform.

Scaling access across critical sectors#

The $1 billion commitment focuses on bringing subsidized access to Daybreak's AI capabilities—including models, security frameworks, training, and technical support—to defenders tasked with protecting essential infrastructure worldwide. In the U.S., Daybreak for America specifically targets systems managing water distribution, electricity, municipal services, and banking. The program will launch a pilot partnership with the Multi-State Information Sharing and Analysis Center (MS-ISAC), which serves cybersecurity needs for state, local, tribal, and territorial governments.

OpenAI also notes it has mobilized utility companies across 40 states and Washington, D.C., to strengthen their cyber defenses using its tools. The initiative builds on Daybreak Defense Network, which now connects over 350 enterprise products and partner services.

Building on recent momentum#

This latest effort follows OpenAI's August release of Daybreak Red and Daybreak Blue—tiered access levels offering different capabilities. Daybreak Red provides researchers and advanced defenders access to GPT-5.6 Cyber for sophisticated work like discovering zero-day vulnerabilities and constructing exploit chains. Daybreak Blue grants approved security professionals GPT-5.6 Sol for code review, malware analysis, incident response, and vulnerability discovery.

The push reflects growing urgency around cyber threats to essential services. In July, cyber actors targeted U.S. water and wastewater systems, prompting OpenAI to respond by offering affected states and utilities $1 million in free API credits, Daybreak access, and hands-on technical help to review configurations, validate security findings, and deploy patches without service interruptions.

Competitive pressure and rising stakes#

OpenAI is not alone in this space. Anthropic has launched Glasswing, a similar consortium-based initiative using Claude Mythos Preview, which has achieved comparable benchmarks and even shares some industry partners. Both platforms aim to place advanced AI capabilities into the hands of defenders—though through different technical approaches and deployment models.