SignalSafety practiceSG-0040
OpenAI opened GPT-Rosalind, its gated biology model, to US university labs at no cost
Labs can get up to 30 seats for up to a year, but only after OpenAI vets the lab lead, the institution and the intended research. Approval also names an institutional safety contact, and OpenAI keeps monitoring use afterwards, with its safety checks adjusted to the approved research for registered high-sensitivity work. The lab lead decides which members get access and answers for their use. The listed research areas include vaccines against emerging viral variants and how pathogens interact with their hosts.
Why it matters
OpenAI gates GPT-Rosalind because the same biology skills that speed up research could help someone misuse it. Opening it to academic labs means safety depends less on who the customer is and more on how well OpenAI vets and monitors individual researchers. OpenAI vets the lab lead and the lab, and the lead then names up to 30 members, so that one approval carries most of the weight.
Sources
Read the reporting
1 source. Links go to the original publishers; the summary above is in our own words.
- primaryBringing GPT-Rosalind to academic research labsOpenAI · Oct. 15, 2026openai.com/index/introducing-the-rosalind-academic-plan/
Signals
More from the news desk
Cite and share
Use this signal
Citation
Paperclip Index. “OpenAI opened GPT-Rosalind, its gated biology model, to US university labs at no cost.” Signal SG-0040. Reported Oct. 15, 2026; updated Oct. 15, 2026. https://paperclipindex.com/signal/SG-0040