OpenAI Researcher Miles Wang Is in Talks to Launch a $2 Billion AI Drug Discovery Startup

OpenAI researcher Miles Wang is leaving the ChatGPT maker to start an AI drug discovery company, with fundraising talks reportedly centering on a $200 million round at a $2 billion valuation. The move matters because it underscores how quickly investors are pouring money into artificial intelligence tools aimed at speeding up pharmaceutical research.

According to multiple people familiar with the discussions, several other OpenAI researchers are expected to join Wang at the new venture. The startup is said to be focused on building AI models for drug discovery, an area that has become one of the hottest corners of the AI market as founders, scientists and venture capital firms chase faster ways to identify treatments and repurpose existing medicines.

Wang, who joined OpenAI in 2024 after leaving Harvard, has worked on research that uses AI to accelerate scientific and biological discovery. His departure would add another high-profile name to a growing list of OpenAI alumni moving into life sciences startups, where the potential upside is enormous and the timelines can be shorter than in traditional drug development.

The reported deal is still in flux. The sources who described the financing said the talks are ongoing, the numbers could change and no final agreement has been completed. Wang also disputed the reported funding figures and the description of the company, though he did not provide an alternative set of details. Lightspeed, which is reportedly in discussions to lead the round, did not respond to a request for comment.

Even with those caveats, the size of the rumored raise would place the new startup among the most heavily backed AI-for-science ventures at a time when investor appetite for drug-discovery platforms has surged. Just days before the report emerged, another startup in the same broad category, Chai Discovery, announced a major financing. In May, Isomorphic Labs, the Google DeepMind spinout focused on drug discovery, also landed a multibillion-dollar round. Together, those deals show that the race to build AI systems for chemistry and biology is moving from experiment to strategic priority.

Why investors are betting big on AI drug discovery

AI drug discovery has become attractive for a simple reason: if software can shorten the process of finding promising molecules, predicting interactions or identifying new uses for existing drugs, it could save years of work and millions of dollars. That promise is especially compelling in pharmaceuticals, where traditional development is slow, expensive and full of dead ends.

Unlike many consumer AI products, drug-discovery tools have a clear business case tied to measurable scientific outcomes. If a model can help researchers prioritize targets, propose candidate compounds or revisit shelved drugs that failed for non-safety reasons, a company can potentially generate value much faster than a startup building entirely new molecules from scratch.

That speed matters. Existing FDA-approved drugs already have safety data behind them, which can make repurposing programs more efficient than inventing a therapy from the ground up. Some of the people familiar with Wang’s plans said his startup may be aiming not only at new drug candidates, but also at finding new applications for medicines already on the market or drugs that were previously abandoned in trials.

How repurposing drugs can accelerate revenue

Repurposing can move faster because the scientific and regulatory groundwork is partially done. When a medicine has already cleared safety hurdles, developers may be able to focus on efficacy in a new indication, which can reduce risk and compress the path to clinical and commercial progress.

That does not make the work easy. Repurposed drugs still need robust evidence, clinical trials and regulatory clearance for new uses. But the starting point is often more favorable than a completely novel compound, making the field attractive to investors who want both scientific ambition and a more practical route to returns.

• Existing safety data can lower early-stage risk.
• Known compounds may move faster through development pipelines.
• Successful repurposing can create earlier commercial opportunities.

Who is Miles Wang and why does his move matter?

Miles Wang is a young OpenAI researcher whose work has centered on using artificial intelligence to speed up scientific discovery. He joined OpenAI in 2024 after leaving Harvard, where he was studying computer science but had not completed his degree.

His background reflects a broader pattern in Silicon Valley and on Sand Hill Road: investors are once again comfortable backing early-career founders, including those who left elite universities before graduation. That trend has become especially visible in AI, where technical talent and research credentials can be valued as highly as operating experience.

At OpenAI, Wang co-authored research examining how AI models might automate and accelerate scientific discovery. That work likely made him a fit for the new startup he is now planning, particularly if the company is aiming to build foundation models or specialized systems for biology and chemistry.

Sources familiar with Wang’s plans say his new venture is intended to apply AI directly to drug discovery, with several OpenAI researchers expected to join him.

If that happens, the company could launch with an unusually strong research pedigree. Teams formed from top AI labs have become a common source of new startups, especially when founders want to translate general-purpose model expertise into a narrow but high-value industry.

What would the reported funding mean?

If the financing comes together at the reported size, the new company would enter the market with enormous investor expectations from day one. A $200 million round at a $2 billion valuation would signal that backers believe the startup can compete with some of the best-funded names in AI-driven life sciences.

Those numbers are not yet finalized, and the sources stressed that the deal terms may still shift. But even the possibility of a raise at that level reveals how much capital is chasing the intersection of AI and biotech.

For venture firms, the appeal is straightforward. AI drug-discovery companies promise large addressable markets, potential platform economics and the chance to influence an industry that spends heavily on R&D. For founders, the opportunity is to use advanced models to solve problems that have resisted brute-force lab work for decades.

How does this fit into the broader AI race in life sciences?

It fits into a fast-moving wave of startup activity that has turned drug discovery into one of the most competitive AI battlegrounds. The field is being shaped by a mix of Big Tech spinouts, research-heavy startups and founders with direct experience inside frontier AI labs.

Chai Discovery is one recent example. The company, founded two years ago, says its models can predict molecular interactions to help identify new drugs. Its announcement of a $400 million round at a $3.8 billion valuation showed how quickly investors are rewarding progress in this niche.

Another major signal came from Isomorphic Labs, the DeepMind spinout that raised a $2.1 billion Series B in May. That deal reinforced the idea that the market sees AI drug discovery not as a speculative side bet but as a platform category with the potential to reshape a major industry.

In that context, Wang’s planned launch looks less like an isolated founder move and more like part of an emerging pipeline from frontier AI research into applied biotech. Researchers who have trained or worked on advanced models often leave to build companies that target specific scientific bottlenecks, and drug discovery is among the most lucrative of those targets.

Why OpenAI alumni are showing up in biotech startups

OpenAI alumni have become especially visible in adjacent categories because the company sits at the center of the current AI wave. Researchers who have worked on large models often leave with a deep understanding of scaling, evaluation and model behavior, skills that transfer well to scientific applications.

Drug discovery is a natural destination for that talent because it demands pattern recognition, data handling and the ability to make predictions in complex, noisy environments. Those are precisely the kinds of problems that modern AI systems are increasingly designed to tackle.

• Frontier AI labs train researchers on cutting-edge model development.
• Biology and chemistry generate large, complex datasets.
• Drug discovery offers a high-value commercial use case for AI.

What we know — and what remains unconfirmed

Here is the clearest picture available so far: Wang is preparing to depart OpenAI, he plans to build an AI drug discovery company, and he may be joined by other OpenAI researchers. The startup is reportedly in fundraising talks, but the specific amount and valuation have not been locked in.

There are also open questions about the company’s product strategy. Some sources say it may focus on models for discovering new drugs, while others suggest the company could emphasize repurposing existing medicines and previously failed compounds. In practice, those approaches are not mutually exclusive, and many drug-discovery platforms try to address both early-stage and later-stage opportunities.

Wang’s public response adds another layer of uncertainty. By disputing the reported figures and description, he signaled that at least some details being circulated are inaccurate. Without a finalized announcement from the company or its backers, the exact structure of the startup remains unconfirmed.

Why the timing is notable

The timing of the reported launch is important because the AI market is entering a more selective phase. Investors have become more focused on applications that can point to real-world value rather than general hype, and life sciences fits that brief well.

Drug discovery combines high technical difficulty with potentially massive payoff. That combination is exactly what venture investors often seek, especially when the founding team comes from a top-tier AI lab and the startup’s pitch is grounded in a clear industrial pain point.

At the same time, the field remains risky. Scientific progress can be slow, model performance can lag behind expectations and biology is harder to predict than many software investors initially assume. Still, the money flowing into this area suggests that the market is willing to keep funding the experiment.

Timeline of key developments

The events below show how quickly the story has unfolded across OpenAI, biotech and venture capital.

What happens next?

The next milestone is whether Wang’s fundraising actually closes and whether more OpenAI researchers do, in fact, join him. If the company secures a large round, it would likely move quickly to recruit talent, define a product strategy and build its scientific platform.

Investors will also be watching whether the startup positions itself as a general AI model company for pharma or as a more targeted tool for one part of the drug-development pipeline. Those strategic choices will shape both the startup’s customer base and how quickly it can show results.

For now, the report adds to a broader pattern: top AI researchers are increasingly leaving major labs to pursue narrowly focused, capital-intensive businesses in science and medicine. In that sense, Wang’s planned departure is not just a personnel change at OpenAI. It is another sign that the race to use AI in drug discovery is intensifying fast.

And if the reported funding terms prove accurate, the startup will begin life with the kind of valuation that reflects both the promise and the pressure of that race.