OpenAI가 GPT-Rosalind를 공개하면서 AI가 신약 발견을 방해합니다 - kmjournal.net

The drug discovery industry is feeling the shock. OpenAI has introduced GPT-Rosalind, a new AI model built specifically for life sciences research, and the market reacted almost immediately. Stocks tied to traditional drug development dropped, signaling a deeper concern. This is not just about efficiency. It is about a shift in how new medicines are created. Aiming to Cut Years Off Drug Development OpenAI announced GPT-Rosalind on April 16, positioning it as more than a data analysis tool. The model is designed to support the entire research workflow, from reviewing scientific literature to forming hypotheses, designing experiments, and interpreting results. In practice, that means tasks that once took researchers months can now be assisted step by step by AI. Drug development typically takes 10 to 15 years, with early-stage target discovery being one of the most time-consuming and expensive phases. GPT-Rosalind focuses on reducing this bottleneck, where much of the uncertainty and cost are concentrated. Near-Expert Performance in Key Tests Early benchmarks suggest the model is already performing at a high level. In LABBench2, a widely used evaluation for life science AI systems, GPT-Rosalind outperformed GPT-5.4 in 6 out of 11 tasks. In collaboration tests with Dyno Therapeutics, the model showed strong results in RNA sequence prediction, reaching a level comparable to the top 5 percent of human experts. In some areas, it is already approaching expert-level capability. OpenAI, however, makes one thing clear. The model is not replacing scientists. Human judgment and validation remain essential. Big Pharma Moves First GPT-Rosalind is currently available as a research preview through API access. It is being tested by a limited group of partners, including major pharmaceutical companies like Amgen and Moderna, as well as institutions such as the Allen Institute and Thermo Fisher Scientific. This is not just a tech demo. It is already being integrated into real-world drug development pipelines. That shift matters. AI is no longer just helping write papers or analyze datasets. It is starting to play a role inside the core process of developing new drugs. Market Reacts to Potential Disruption The strongest signal came from the market. Right after the announcement, stocks related to drug development declined. IQVIA Holdings dropped as much as 3.2 percent במהלך trading. Charles River Laboratories fell 2.6 percent. AI-focused biotech firms like Recursion Pharmaceuticals and Schrödinger also saw declines of more than 5 percent. Investors appear to be pricing in a new reality. If AI can take over parts of the drug discovery pipeline, companies that rely on traditional contract research or platform-based models could face pressure. A Turning Point for AI in Biotech GPT-Rosalind suggests a bigger shift is underway. AI is moving beyond support roles and into decision-making layers within scientific research. The technology is still evolving, and full automation of drug discovery is not here yet. But the direction is clear. The timeline, cost structure, and even the competitive landscape of the pharmaceutical industry may not look the same for much longer. by Ju-baek Shinㅣ[email protected]