Robots Atlas>ROBOTS ATLAS
Artificial Intelligence

Ex-OpenAI researcher raises $200M for AI drug discovery at $2B valuation

Ex-OpenAI researcher raises $200M for AI drug discovery at $2B valuation

Miles Wang, an OpenAI researcher who specialized in accelerating biological discovery with AI, is leaving the company to launch a drug discovery startup at a $2B valuation. On the same day — July 14, 2026 — Chai Discovery, another OpenAI-rooted startup, announced a $400M raise at a $3.8B valuation.

Key takeaways

  • Wang is in talks to raise approximately $200M at a $2B valuation, with Lightspeed in discussions to lead
  • Wang joined OpenAI in 2024 after dropping out of Harvard's computer science program
  • Chai Discovery raised $400M at a $3.8B valuation on the same day
  • Isomorphic Labs (Google DeepMind spinout) raised a $2.1B Series B in May 2026
  • Wang's startup may focus on Drug repurposing: A strategy of finding new therapeutic uses for molecules already approved by the FDA — it shortens the path to market because the safety profile is already established. — finding new uses for FDA-approved molecules

Who is Miles Wang

Wang joined OpenAI in 2024 after leaving Harvard, where he was pursuing a bachelor's in computer science. At OpenAI, he co-authored research papers evaluating how AI models can automate and accelerate scientific discovery — including a study on AI-driven acceleration of Wet lab: An experimental biology laboratory where physical experiments with chemical and biological substances are conducted — as opposed to a dry lab, which operates purely on data and computation. biological research. Wang disputed TechCrunch's reported funding figures and company description but did not provide his own numbers. Lightspeed did not respond to a request for comment.

Drug repurposing as a business model

Several sources familiar with Wang's plans say the startup may focus on drug repurposing — finding new therapeutic indications for molecules already approved by the FDA or those that failed clinical trials for unrelated reasons. The advantage is a substantially shorter time to revenue: drugs with established safety profiles bypass the most expensive and risky phase of clinical testing.

OpenAI itself has explored adjacent territory. In 2025, the company published research showing how AI models can accelerate wet lab biology. Wang was a co-author on that work.

A moment for AI drug discovery

Chai Discovery raised $400M at a $3.8B valuation from Coatue Management. Co-founder Josh Meier, like Wang, has OpenAI experience. Isomorphic Labs, a Google DeepMind spinout developing AI models for drug discovery, closed a $2.1B Series B in May 2026 — one of the largest rounds in the history of this category.

The OpenAI spin-out pattern

Wang and Meier are part of a clear pattern: senior OpenAI researchers launching independent companies applying AI to scientific and biomedical problems. For OpenAI, this is simultaneously a challenge and a validation. A challenge — because it loses key researchers at a moment of intense competition with Anthropic, Google DeepMind and Meta AI. A validation — because it confirms that the firm has been a training ground for a generation of builders.

Why this matters

The convergence — Wang, Chai Discovery, and Isomorphic Labs in a short span — creates a picture of a category in which investors are prepared to assign billion-dollar valuations ahead of confirmed clinical products. AI drug discovery is a bet that models capable of generating and verifying biochemical hypotheses will compress the decade-plus cycle from discovery to approval. The field is young and no FDA-approved drug can yet be attributed solely to an AI model. Current valuations reflect expectations, not outcomes.

What's next

  • Wang and his investors must finalize the round — details could change
  • Chai Discovery announced plans to expand its team and accelerate work on molecular interaction models after closing its $400M round
  • Isomorphic Labs is in advanced talks with pharmaceutical companies about clinical partnerships

Sources

Share this article