OpenAI researcher Miles Wang launches AI drug discovery startup at $2B valuation

An OpenAI researcher, Miles Wang, is leaving to launch a startup focused on AI models for drug discovery, reportedly in talks to raise $200 million at a $2 billion valuation. The funding discussions, with Lightspeed reportedly leading, highlight growing investor appetite for applying foundation-model-style AI to biology—especially drug repurposing. Wang’s departure also pulls several other OpenAI researchers along, signaling that top AI talent sees drug discovery as a high-leverage application for large language models and molecular prediction.

Wang’s approach is reportedly centered on finding new uses for existing drugs, including those that failed in clinical trials, rather than designing entirely new molecules. This is a deliberate strategy: repurposing FDA-approved drugs bypasses early safety testing, slashing time and cost to revenue. The move mirrors the trajectory of Chai Discovery, another OpenAI-connected startup that just raised $400M at a $3.8B valuation for molecular interaction prediction, and Isomorphic Labs, the DeepMind spinout that secured a $2.1B Series B in May. All three bets depend on whether AI can reliably model complex biochemistry.

For technical readers, the key tension is between the hype and the hard science. Drug discovery is notorious for high failure rates, and no amount of compute guarantees a working molecule. What makes this worth watching is the specific focus on repurposing—a narrower, more tractable problem that could generate real revenue faster. If Wang’s models can reliably match existing drugs to new diseases, the payoff is enormous; if not, it’s another cautionary tale about AI in biotech.