SecureBio lands $17.2M OpenAI Foundation grant for pandemic early warning

SecureBio, a biosecurity research organization, has accepted a $17.2M grant from the OpenAI Foundation (OAIF), the nonprofit arm founded by OpenAI cofounder Wojciech Zaremba and national security policy lead Yo Shavit. The grant will expand SecureBio's outbreak early warning system, which uses metagenomic sequencing to detect any DNA or RNA in wastewater and nasal swab samples without predefined targets—a technique capable of catching engineered pathogens before they spread widely. The partnership emerged from conversations about how AI models could increasingly help bad actors create pandemics, and whether frontier labs like OpenAI should slow biological capability work.

SecureBio's founder, however, had two major concerns: independence and safety incentives. To address them, the grant is strictly earmarked for SecureBio's Detection division, while its AI evaluation division—which assesses models from multiple firms, including OpenAI—remains fully separate. OAIF gained no formal control over SecureBio's public statements or model critiques, and actually asked to review SecureBio's policies to ensure they were robust. The founder acknowledged that incentives still matter, but called the grant "key" and "incredibly exciting" for accelerating turnaround on outbreak detection, ultimately reducing the risk of a silent engineered pandemic.