OpenAI researcher Miles Wang is leaving to form an AI drug-discovery startup, according to TechCrunch, but the reported financing remains disputed.

The outlet cited people familiar with the plans as saying the company is in talks to raise about $200 million at a $2 billion valuation.

The reported funding remains unresolved after Wang challenged TechCrunch's figures and company description.

The outlet said he did not provide corrected numbers or alternative company details, and Lightspeed did not respond to a request for comment.

$200 Million Funding Talks Remain Unfinished

TechCrunch reported that Lightspeed is in discussions to lead the round.

The talks are still ongoing, the deal may not close and terms could change, according to the outlet's source account.

The planned company would focus on AI models for drug discovery.

Several other OpenAI researchers are expected to join, according to TechCrunch, which described Wang's OpenAI work as including AI for scientific and biological discovery.

People cited by TechCrunch said the startup may work on models that identify new uses for existing drugs and possibly medicines that previously failed in trials.

TechCrunch noted that repurposing FDA-approved drugs can shorten the time to revenue compared with developing new drugs from scratch because those medicines have already been tested for safety.

AI Drug Discovery Funding Has Larger Comparables

The reported financing talks come alongside larger AI drug-discovery rounds.

TechCrunch cited Chai Discovery's Tuesday announcement of a $400 million raise at a $3.8 billion valuation, and noted that co-founder Josh Meier had also worked at OpenAI.

The same article cited Google DeepMind spinout Isomorphic Labs, which develops AI models for drug discovery and raised a $2.1 billion Series B in May.

Those figures provide public comparables for investor interest, but they do not confirm Wang's funding terms.

OpenAI Research Work Provides The Technical Link

Wang joined OpenAI in 2024 after leaving Harvard, where he had been working on a computer science degree, according to TechCrunch.

At OpenAI, he co-authored papers that evaluated how AI models can automate and accelerate scientific discovery.

That research background gives the startup a source-backed AI model focus rather than a general biotech financing angle.

TechCrunch did not name final co-founders, disclose signed investors, identify drug-development partners, provide corrected funding figures or give a launch date for the new company.