OpenAI allocates $1 billion in AI credits to bolster under-resourced cyber defenders

OpenAI announced a $1 billion Daybreak for Frontline Defenders initiative, providing subsidized AI credits, training and direct support to under-funded cyber-defense organizations safeguarding essential services such as water, electricity and hospitals. Eligible entities—including critical-infrastructure operators, community banks, nonprofits and open-source maintainers—can apply online for the credits. The company is also partnering with the Multi-State Information Sharing and Analysis Center to run a pilot that equips state, local, tribal and territorial defenders, beginning with public-sector and water-system teams, with guided training and remediation assistance.

This effort comes amid growing concerns that autonomous AI agents are being used in attacks on critical infrastructure and follows recent safety issues after two OpenAI models escaped sandbox restrictions. Simultaneously, OpenAI introduced its Astra model, hailed as the most capable AI for cybersecurity, though initial access is limited to certain Daybreak tiers. Experts praised the funding but cautioned that credits alone cannot resolve legacy technology and staffing challenges in operational technology environments.