# ByteDance spins out Anew Labs: when AI companies start owning drug assets

> ByteDance spun out its AI drug discovery unit as Anew Labs, keeping 56% of a $1.5bn company with its own preclinical pipeline. The structure shows how far large AI companies now go into drug development, and why they separate it from the core business.

URL: https://www.ch-healthtech.com/insights/bytedance-anew-labs-spinout-ai-companies-owning-drug-assets
Markdown: https://www.ch-healthtech.com/insights/bytedance-anew-labs-spinout-ai-companies-owning-drug-assets.md
Published: 2026-09-18
Author: Christian Hein
Tags: industry/tech-bio, technology/artificial-intelligence, activity/venture-investment, function/pre-clinical-research, industry/biotech, activity/series-a-seed-funding

---


## TL;DR

ByteDance has spun out Anew Labs, its AI drug discovery unit, which just raised $290m at a $1.5bn valuation, keeping 56% ownership. I think this structure, building the AI capability inside a tech company then spinning it into a separate biotech, is one we may see more. Anew's pipeline is still preclinical, so the $1.5bn valuation tells us very little about eventual clinical success.

TikTok's parent company now owns 56% of a $1.5bn AI drug discovery company.
ByteDance has spun out Anew Labs, its AI drug discovery operation, which has just raised $290m from HSG, IDG Capital, GL Ventures, the venture arm of Hillhouse Investment, and others.

According to one of Reuters' sources, ByteDance spun the unit out because AI drug discovery follows a different "industry logic and management approach" from its core business. That is a fairly important admission.

## Where AI companies stop in life sciences

We are seeing more large AI companies push into life sciences, and I keep coming back to the same question: where do they eventually stop? At the model layer? At the discovery platform? Or do they start owning drug assets themselves?

Anew has already gone quite a long way down that road. It has its own AI stack spanning structure prediction, molecular design and scientific reasoning, alongside experimental science. Its website lists four pipeline programmes, including an IL-17 programme where it has presented AI-designed small molecules. So this is already well beyond providing software to pharma.

The structure ByteDance has chosen is worth watching. Build the capability inside a technology company, then put it into a separate biotech with its own capital, management and development model while retaining the majority of the economics. I suspect we may see more of this.

Large technology companies have obvious advantages in models, compute and technical talent. Drug development brings very different capital cycles, experimental infrastructure, regulation and clinical risk. A separate company gives those two worlds more room to operate on their own terms.

## Why a separate company makes sense

Anew is still early and its disclosed pipeline remains preclinical. A $1.5bn valuation tells us very little about eventual clinical success.

We already know China as a hotspot for fast early-stage biotech development. Add the AI scale of a company like ByteDance to that ecosystem and this is worth watching.