OPENAI JUST GAVE BIG PHARMA A SMARTER DRUG DISCOVERY AI — AND NOVO NORDISK GOT FIRST ACCESS OpenAI has released an updated version of GPT-Rosalind, its specialized artificial intelligence model for life sciences research, combining the advanced reasoning and agentic coding capabilities of GPT-5.5 with deep domain intelligence covering drug discovery, genomics, medicinal chemistry, and protein engineering. The upgrade delivers specific numbers worth paying attention to. The updated model uses 31 percent fewer tokens than GPT-5.5 while improving accuracy on drug discovery benchmarks. It scores 0.751 on BixBench, a key evaluation for biochemistry reasoning. OpenAI describes it as built for workflows spanning target identification, lead optimization, and experimental design. Novo Nordisk was given early access to the updated model, and the broader rollout is targeting qualified enterprise customers through a trusted access program. Other partners already working with GPT-Rosalind include Amgen, Moderna, the Allen Institute, and Thermo Fisher Scientific. OpenAI named the model after Rosalind Franklin, the crystallographer whose painstaking X-ray work helped reveal the structure of DNA — and whose contributions were famously minimized by the men who received the Nobel Prize for the discovery. The ambition is straightforward: compress the drug discovery timeline from years to months using AI that can reason across the entire research stack at once. Keywords: OpenAI GPT-Rosalind, drug discovery AI, Novo Nordisk AI, life sciences AI model