A "virtual biotech company" run entirely by software is no longer a thought experiment: researchers have deployed 37,000 AI agents as a simulated drug-development firm, and the digital organization went beyond imitating the daily workflow of a real biotech to propose concrete treatment strategies for lung cancer, according to a report whose central claims were checked against eight primary sources.
The project stands out for its scale. Rather than a handful of AI assistants, the virtual firm fields 37,000 AI agents, organized to replicate the functions of a working biotech company โ from early research through development. The result is less a chatbot demonstration than a simulated organization, one in which autonomous agents carry out the kind of coordinated, specialized labor that human teams normally provide. The report describes the entity as a functioning "virtual biotech company," a designation that captures both its structure and its ambition.
What elevates the work above a modeling exercise is its output. The AI-driven firm did not merely mimic processes; it generated proposals for lung cancer treatment strategies, moving from organizational simulation to medically relevant recommendations. Verification of the findings, conducted through a review of eight primary sources, confirmed the claim as fact. The convergence of agentic AI โ systems that act and collaborate rather than simply answer โ with drug development has been one of the most closely watched frontiers in the field, and this project offers one of the clearest demonstrations yet of agents operating at company scale on a disease as complex as cancer.
For an industry betting that agentic systems can compress the time and cost of drug discovery, a 37,000-agent virtual company proposes a template: simulate the institution, not just the molecule. The approach could reshape how early-stage research is staffed and sequenced, with AI agents handling parallel lines of investigation that human teams pursue sequentially. It also signals where competitive advantage may shift in the AI-bio convergence race โ toward those able to orchestrate very large agent populations around concrete disease targets.
Based on the verified record, the claim that a virtual biotech company powered by 37,000 AI agents has emerged and proposed lung cancer treatment strategies is True.