\n\n
A new OpenAI life‑sciences model, gpt‑rosalind‑discovery, has quietly appeared on the API pricing page, signaling the company’s deeper push into vertical AI markets. The model, launched on October 5 for trusted‑access partners, targets medicinal chemistry and genomics, mirroring OpenAI’s strategy to supply specialized tools rather than conduct its own drug programs.
Key developments include the maturation of AI‑assisted drug discovery across six core tasks—target identification, virtual screening, generative design, ADMET prediction, repurposing and patient stratification. The most concrete clinical proof to date is the generative‑AI‑discovered TNIK inhibitor rentosertib, which completed a phase 2a safety trial of 71 patients (June 2025) and entered a 52‑week phase III study in China (September 2026). While the trial’s safety endpoint is modest, it represents the first phase III of a generative‑AI‑driven drug.
Regulators are beginning to formalise AI oversight. FDA’s draft guidance (January 2025) and joint FDA‑EMA principles (January 2026) require a defined context of use and risk‑based credibility assessments, underscoring that AI outputs must be well‑governed before supporting submissions.
Commercially, the market is expanding rapidly. Grand View Research projects AI drug‑discovery spending at $2.9 billion in 2026, climbing to $13.8 billion by 2033, while McKinsey notes investment doubled to $8.4 billion in 2025. Start‑ups such as BioinvestGPT are demonstrating predictive power by accurately simulating trial outcomes—e.g., correctly forecasting the failure of Novartis’ del‑desiran.
Why it matters: Specialized models like gpt‑rosalind‑discovery lower the cost of early‑stage discovery, potentially improving hit‑to‑lead filtering and shortening timelines. Successful AI‑driven trials would validate the technology’s promise beyond in‑silico gains, influencing funding, partnership strategies, and regulatory frameworks.
What to watch next: (1) Results from the rentosertib phase III trial, which will be a decisive test of generative‑AI efficacy; (2) Adoption of OpenAI’s Rosalind model by major pharma partners such as Amgen and Moderna; (3) Implementation of FDA/EMA AI guidance in regulatory submissions; and (4) Continued market growth as AI tool spend accelerates toward the projected $13.8 billion horizon. [[1, 2]]