Rising AI attack risks prompt OpenAI to allocate $1 billion to shore up defenses for its critical infrastructure.

Beating AI News (from Insight) – OpenAI has launched "Daybreak for Frontline Defenders", a $1 billion initiative to provide AI-powered cybersecurity defense for critical infrastructure. Rather than direct grants to organizations, the funding will be delivered as Daybreak model access, training, and technical support. The initial rollout targets U.S. water utilities, power grids, local governments, community banks, nonprofits, and open-source projects, with deployment planned over the next six months. Daybreak helps these entities identify vulnerabilities, analyze anomalies, validate risks, and test patches. OpenAI stated that approximately 2,000 organizations and workspaces are already using the tool, with plans to integrate more than 35 additional enterprise products and services. The initiative was announced alongside GPT-6 Astra, OpenAI’s first model to meet critical cybersecurity capability thresholds. When paired with supporting tools, Astra can already locate unknown vulnerabilities and research new exploitation methods without step-by-step human guidance. OpenAI forecasts that AI-driven cyberattacks will grow more widespread and sophisticated in the coming months.