AI Risk Clock
Doomsday Clock2 min to midnight
→
⚖️ Neutral edit
axios.com2026-05-29

Exclusive: OpenAI launches biodefense program - Axios

ModelsSafety

OpenAI has launched the Rosalind Biodefense Program, a new initiative to develop biodefense and pandemic preparedness capabilities using artificial intelligence. The program will offer a specialized model called GPT-Rosalind for life sciences research to 'trusted developers' who are operationalizing biodefense tools. The announcement was made via an exclusive report by Axios.

The GPT-Rosalind model is designed to assist in research related to biological threats and pandemic response. OpenAI stated that the model will be made available to a select group of developers who have been vetted for trustworthiness, though the specific criteria for selection were not detailed in the initial announcement. OpenAI emphasized that the program is part of its ongoing commitment to responsible AI development, focusing on beneficial use cases while mitigating potential harms.

The initiative marks OpenAI's entry into the field of AI for biodefense, an area where artificial intelligence can accelerate research but also raises concerns about dual-use risks. By limiting access to trusted developers, OpenAI aims to balance the potential benefits with safety considerations. The program is separate from OpenAI's general API offerings and represents a targeted application of its technology for a specific, safety-critical domain.

Read this story in another voice
● REC · 2026