Unfamiliar Newcomer in the AI Pharmaceutical Industry

Anew Labs is not an isolated case of large internet enterprises entering the life sciences sector. Over the past few years, Baidu Group, Tencent Holdings, Alibaba, Huawei and other companies have successively applied their AI capabilities to drug R&D, with some building in-house teams, some investing in industrial firms, and some attempting to leverage large models to enter the field of drug discovery. In this cross-sector expansion, ByteDance has gone a step further, developing not only AI models but also drug pipelines.

In September 2026, an AI pharmaceutical company spun out from ByteDance secured USD 290 million in angel round financing, making it the first AI pharmaceutical company in China to reach a valuation of RMB 10 billion at its initial financing round.

The company, named Anew Labs, had only been independent from ByteDance for 3 months, with no mature products and its drug pipelines still in the preclinical stage, yet it drew a large number of star institutional investors competing to participate in the financing, which made many AI pharmaceutical startups extremely envious.

Li Mi, an investor participating in this round of financing, told the Economic Observer that the competition for this financing was extremely fierce, and he "grabbed" the investment quota from more than 100 institutions.

For ByteDance, this is a rare business spin-off, and the last time it received such high attention dates back to the spin-off of Dongchedi at the end of 2023. Before the spin-off, Anew Labs had operated within ByteDance for 5 years, and ByteDance remains the controlling shareholder after this spin-off.

Anew Labs is not an isolated case of large internet enterprises entering the life sciences sector. Over the past few years, Baidu Group (BIDU.NASDAQ/09888.HK), Tencent Holdings (00700.HK), Alibaba (BABA.NASDAQ/09988.HK), Huawei and other companies have successively applied their AI capabilities to drug R&D, with some building in-house teams, some investing in industrial firms, and some attempting to leverage large models to enter the field of drug discovery. In this cross-sector expansion, ByteDance has gone a step further, developing not only AI models but also drug pipelines.

Behind the capital's enthusiasm lies an unproven cross-sector experiment: can ByteDance's computing power, talent pool and capital be truly transformed into pharmaceutical R&D capabilities? As Anew Labs shifts from model development to drug development, can it build its own core competitiveness in an R&D system completely different from that of the internet industry?

The Economic Observer sent a letter to Anew Labs regarding issues including the reasons for the spin-off, the significance of the financing and its future development strategy, and no official response had been received as of press time. Liu Kai, the head of Anew Labs, responded that the company hopes to focus on R&D at the current stage and will not accept interviews for the time being.

Unfamiliar Visitor

In April 2026, at the Annual Meeting of the American Association of Immunologists, ByteDance's Anew Labs disclosed data of a small-molecule candidate drug targeting IL-17 for the first time, which is the core pipeline of its AI pharmaceutical business.

In this venue that has long been dominated by global biomedical enterprises, ByteDance is like an unfamiliar visitor. This Chinese company famous for its internet and AI businesses came for the first time with its candidate drug designed by AI.

Liu Kai, head of Anew Labs, later reposted relevant news on his WeChat Moments and wrote: "I hope we can use AI to make something in the pharmaceutical industry that truly promotes the progress of the whole sector. Though it is extremely difficult, I think this is where the real value of AI lies."

In the first few years of its layout in AI pharmaceuticals, ByteDance kept a very low profile. Team members rarely participated in pharmaceutical industry forums, and seldom took the initiative to promote after publishing papers, so that some people in the industry once thought this was just a casual trial by a large tech company. Zhou Yiming, founder of Jingjing Biotech, did not seriously try its algorithm until he accidentally saw a paper published by Anew Labs.

"It is excellent and elegant." That was his first impression after using it.

The list of investors in Anew Labs' initial financing is very star-studded. Well-known investment institutions including HSG, IDG Capital, GL Ventures, 5Y Capital, Gaorong Ventures, Chunhua Capital, Boyu Capital participated in the investment, while China Biopharmaceuticals (01177.HK) and Shanghai Future Industry Fund joined as strategic investors.

"Anew Labs has achieved the current results in just 5 years, which is very fast. If their efficiency was not high, we would not participate in the investment," said Li Mi.

Li Mi has long focused on the AI pharmaceutical and brain-computer interface tracks. He found that the two tracks are equally popular, but the financing scenarios are very different. Many brain-computer interface projects are chased by investors, but there are very few AI pharmaceutical projects that can attract concentrated bets from institutions, and Anew Labs is exactly one of them.

The huge amount of financing made many peers feel both envious and surprised. Zhao Yu, co-founder of AI pharmaceutical company Zeyuan Tech, said frankly that he "could not understand" the situation, and believed that the aura of the large tech company accounted for a large proportion.

"Many people still do not fully understand AI pharmaceuticals, and there is no particularly clear answer on how to do it and what level it can reach," Zhou Yiming believes. In a track where the answer is not yet clear, ByteDance's advantages in capital, technology and talent naturally constitute a source of trust.

However, Zhou Yiming is not surprised by Anew Labs' current high profile.

As early as 2024, Jingjing Biotech began to use Anew Labs' protein model, making it one of the earliest AI pharmaceutical companies to access its technology. In Zhou Yiming's view, ByteDance has very strong algorithm capabilities, which are not inferior to the algorithms of top overseas companies such as AlphaFold and Diffusion. More importantly, these models are completely open source.

At present, Anew Labs has launched biomedical AI models including Protenix and PXDesign, covering drug R&D links such as protein structure prediction and molecular design, and further launched AnewMind, a scientific reasoning large model oriented to drug R&D decision-making.

In Zhou Yiming's view, AI pharmaceutical tools can help in the early stage of drug R&D, but it is difficult to support a large company only by relying on tools. What ultimately determines the value of an AI pharmaceutical company is whether it can continuously develop competitive pipelines.

Anew Labs is also trying to shift from model capabilities to drug pipeline development. At present, it has 4 preclinical pipeline assets, and its core pipeline is a small-molecule candidate drug targeting IL-17, which is in the preclinical lead compound optimization stage. According to Anew Labs' disclosure, this is the world's first full-spectrum oral small molecule inhibitor that targets both IL-17A and IL-17F.

Different from the perspective of AI pharmaceutical practitioners, Li Mi prefers to observe Anew Labs in the global market. In his view, the scale and valuation of Anew Labs' current financing are not high. The financing scale of overseas companies such as Isomorphic Labs and the investment of tech giants such as Microsoft in AI pharmaceuticals have reached the level of tens of billions of dollars.

"If you want to compete with the world's competitive large models, equivalent capital investment is still necessary," Li Mi said.

For industrial capital, behind the valuation of 1.5 billion US dollars, there is another question: can these models eventually be applied in real drug R&D?

Cao Fenze, Assistant Vice President of China Biopharmaceuticals and Head of the Digital Innovation Center, told the Economic Observer that in addition to financial returns, China Biopharmaceuticals hopes to speed up the application of cutting-edge technologies in real pipelines through cooperation with tech giants including ByteDance, discover new drug opportunities with differentiated potential, and form R&D capabilities that can be reused in multiple projects.

In Cao Fenze's view, the advantages of large tech companies and their incubated teams lie in algorithm talents, computing power resources and engineering systems. "We will focus on how these advantages are combined with drug R&D, whether they can produce repeatable and verifiable results in real projects, and whether they have professional teams and investment mechanisms to continuously promote the projects."

When AI Pharmaceuticals Spin Out from Large Tech Companies

Anew Labs is not a team that emerged out of nowhere.

At the end of 2020, ByteDance began to recruit talents in the AI pharmaceutical field. In 2021, Liu Kai led the team to form the AI pharmaceutical team, with about 50 core members, including both AI4S algorithm talents and senior pharmaceutical industry experts. In the following years, this team was hidden inside ByteDance, until June 2026, when it officially broke away from the large tech company system and became the independent Anew Labs.

Liu Kai is not a typical technical entrepreneur. Before joining ByteDance, he worked at IDG Capital and Volcanic Stone Investment successively, in charge of investment business in the medical and biotechnology fields. Now, IDG Capital also appears in the list of investors in Anew Labs' initial financing.

After the spin-off, ByteDance injected the core team, algorithm platform and existing pipeline assets into Anew Labs as a whole, set up its domestic headquarters in Yangpu, Shanghai, and established offices in San Jose, USA and Singapore. ByteDance still maintains controlling stake and continues to provide Volcano Engine computing power support. According to Li Mi's understanding, Anew Labs still maintains a small team of less than 100 people at present.

Why did it choose to spin out from ByteDance at this time?

Li Mi believes that the spin-off of AI pharmaceutical business by large tech companies is first of all a choice of strategic priority. For tech giants, large models and general artificial intelligence are the first battlefield, while AI pharmaceuticals is in the second battlefield. After independence, the team can have greater strategic and execution space.

Zhao Yu sees another layer of logic. AI and pharmaceuticals are two business logics with great differences, and independent operation also facilitates the introduction of external capital. AI pharmaceutical R&D requires a large amount of capital, and its technical maturity has not been fully verified. Introducing external capital can disperse the risks borne by the group itself.

Zhou Yiming believes that internet companies are used to pursuing efficiency and rapid iteration. After a product goes online, feedback can be obtained quickly; but drug R&D is a typical long-term oriented industry. Decisions made today may not see results until one year or even ten years later, accompanied by a high failure rate. This difference determines that the culture and growth logic of the two types of companies are difficult to be compatible.

In addition to ByteDance, large tech companies including Baidu, Tencent, Huawei and Alibaba have also laid out in the biomedical field, but their paths are not the same. At present, Tencent, Alibaba, Huawei and other companies keep their related businesses within the group; ByteDance made its AI pharmaceutical business independent after developing it internally for several years; while BioMap built its team, technology and R&D system in accordance with the logic of life science companies from the very beginning of its establishment.

"These paths reflect different strategic priorities and investment depths of large tech companies. Companies spun out from large tech companies can focus on doing things more efficiently, measure their main business more clearly, and recruit more personnel," Liu Wei, co-founder and CEO of BioMap, analyzed to the Economic Observer.

When founding BioMap, Li Yanhong, founder of Baidu, and Liu Wei did not consider placing the AI pharmaceutical business inside Baidu. Liu Wei said that when life science develops to a certain depth, it will encounter parts that large internet companies are not familiar with. The underlying equipment and data of life science require long-term investment, and it often takes several years to see results. This R&D rhythm is difficult to follow the iteration mode of internet companies.

In Liu Wei's view, the AI capabilities of large tech companies can radiate to thousands of industries. If it is only about AI empowerment, the core technology, solutions, team and culture can share one set of system. If we go deeper into the core problems of life science, we will encounter another set of logic, and we must build a separate system.

Behind the two different paths, there is also a practical problem, that is, how to calculate the return.

Lin Ze, who works at a large investment institution, has long focused on the AI pharmaceutical track. He observed that AI applications such as navigation, speech recognition, and shopping recommendation usually can see quantifiable results within one or two months. If it works well, continue to promote it; if it does not work well, stop it. But the value of AI pharmaceuticals is difficult to calculate in this way. Even if AI helps design a very good molecule, it is difficult to form a clear return mechanism unless the company sells the molecule, or spins out the project or the company for financing. This is an important reason why some large tech companies spin off their AI pharmaceutical business.

Drug Development Cannot Rely Only on Computing Power

Resources and talents are the first obvious advantages ByteDance shows after entering the AI pharmaceutical field.

Li Mi introduced that in the algorithm stage, startups usually can get an investment of 100 million to 200 million US dollars, corresponding to a team of dozens of people and hundred-card level computing power; while ByteDance can invest hundreds of millions of US dollars, use thousand-card level computing power, and have the ability to build a larger team. In addition, ByteDance already has a mature system of AI talents and technologies. The large model team, algorithm talent reserve and the attractiveness of the platform itself can all be migrated to the AI pharmaceutical sector.

However, the advantages in resources cannot be automatically transformed into pharmaceutical R&D capabilities.

In Li Mi's view, AI and biomedical sectors have different discourse systems. For Anew Labs that comes out with the genes of a large tech company, it needs to maintain a sufficiently open mindset to continuously attract top biomedical talents. After recruiting the talents, the original AI-oriented evaluation system also needs to be iterated.

This difference will eventually be reflected in each specific pipeline.

"Providing services for drug development may only earn 1 cent, while drug development itself may earn 9.99 yuan." In Zhou Yiming's view, many tools currently provided by Anew Labs are essentially just a small step to help early drug discovery. It is difficult to support a large company only by making tools, and larger development space needs to be obtained through pipeline development.

Zhou Yiming believes that at the pipeline level, Anew Labs is on the same starting line as other AI pharmaceutical companies and biotechnology companies, and even its starting line is more backward than others. Without biomedical talents who can truly control the pipelines, relying only on algorithm advantages may gradually lose competitiveness.

In the AI pharmaceutical field, different companies have chosen different development paths. One type takes self-developed pipelines as the core and directly participates in drug R&D; the other type focuses on life science large models and platform capabilities to provide underlying AI capabilities for biomedical enterprises. Liu Wei believes that the two paths correspond to different business logics and success conditions. For companies that take self-developed pipelines as the core, whether they can succeed in the end depends on the pipeline project establishment, the selection of targets and indications, as well as luck. If the pipeline is selected wrongly, no matter how good the AI is, it cannot play a role.

After AI pharmaceuticals enters the pipeline stage, the judgment beyond algorithms is particularly important, which is a more unfamiliar challenge for Anew Labs. Choosing what target to target, what molecule to develop, what project to promote, and what project to stop loss, requires people who truly understand pharmaceuticals to make judgments.

When many respondents talked about the importance of biomedical talents, they all mentioned Hua Shen Zhi Yao. In the past year and a half, this company has reached 5 cooperations with domestic and foreign pharmaceutical giants, with a total potential value of more than 7 billion US dollars, which makes global AI pharmaceutical peers envy.

Zhou Yiming believes that the algorithm of Hua Shen Zhi Yao is not outstanding in the industry. One important reason why it can continuously make a number of heavyweight BD (business development) transactions is that it introduced Zhu Zhenping, former R&D president of 3SBio (01530.HK), to be in charge of pipelines.

Zhu Zhenping is a senior new drug R&D scientist, famous for his vision in selecting targets. During his tenure at 3SBio, he once led the R&D of PD-1/VEGF dual antibody. This pipeline was almost "sold at a low price" at one time, but finally was licensed to Pfizer for more than 6 billion US dollars, shocking the whole industry.

In Zhou Yiming's view, if Anew Labs wants to grow bigger and stronger, finding a talent like "Zhu Zhenping" is more important than continuing to increase computing power and expand the algorithm team. But such talents are very scarce and hard to come by.

Anew Labs is also well aware of the importance of biomedical talents. At present, it has formed a scientific advisory team, whose members include many biomedical experts, such as Liu Yongjun, former president of Innovent Biologics, Ma Ji, former chief scientist of Amgen, and Zou Hua, director of protein chemistry science at Takeda California.

Finding people who understand pharmaceuticals is only the first step. How these talents and algorithm talents cooperate is also a test that Anew Labs needs to face. Zhao Yu introduced that these two types of talents naturally have high communication and collaboration costs. Algorithm talents are used to solving problems with clear boundaries, rules and systematic logic, while biomedical talents rely more on the understanding of complex biological phenomena. If the two teams lack chemical reaction, they may fall into internal friction and slow down the business progress.

"Computational chemistry has developed relatively maturely, with deep accumulation