Who hasn't used AI yet?

From personal life to industrial competition, AI is becoming an indispensable keyword in this era.

The pharmaceutical industry is also undergoing the same changes. For innovative pharmaceutical companies, AI has evolved from an "optional" to a "must-have" for exploring future research and development models.

At the first Global Conference on New Drugs in Major Countries (CPIC 2026), this trend was further amplified. From July 22nd to 24th, 20000 participants in the innovative pharmaceutical industry gathered in Shanghai. At this industry event that gathers global innovative pharmaceutical forces, AI pharmaceuticals have become one of the most talked about topics, with multiple related special sessions packed to capacity.

Industry professionals from pharmaceutical companies, investment institutions, and research institutions walked into the scene, hoping to find an answer: Where has AI pharmaceuticals gone? How can AI pharmaceutical companies establish competitive barriers in the face of the accelerated layout of General Big Model Company? How much room for development does AI pharmaceuticals still have?

Although there are different opinions in the industry, a consensus is emerging that AI pharmaceuticals are far from reaching the end, technological evolution and industry exploration are still ongoing, and more new opportunities will be nurtured until more players arrive.

01. Where is AI pharmaceuticals going?

In 2026, the global popularity of AI pharmaceuticals will once again rise.

The capital market is the first to release signals. Overseas AI pharmaceutical company Isomorphic Labs announced the completion of a $2.1 billion Series B financing, setting a new global record for single round AI pharmaceutical financing; The domestic Hong Kong stock market has also gathered the "Three Little Dragons of AI Pharmaceuticals", and the trading amount of BD continues to increase. The entire track has become a focus of attention in the field of innovative drugs.

The industrial sector is also rapidly advancing. From protein structure prediction and molecular design to target discovery and clinical development, AI is constantly deepening the drug development process. According to a research report by Ping An Securities, as of mid June 2026, the number of AI driven clinical pipelines for new drugs has increased from 4 in 2017 to 179, of which 9 have advanced to phase III clinical trials.

Under the wave of enthusiasm, a key question arises: if AI pharmaceuticals are a football game, where has it "kicked"?
This question may seem simple, but it determines the direction of the industry's next stage of development. If the industry is still in its early stages, it means that new technological routes and entrepreneurial opportunities will continue to emerge; If competition enters a mature stage, it means that enterprises will face more intense elimination and differentiation.

At multiple AI pharmaceutical forums held during CPIC 2026, guests from AI pharmaceutical companies, computational biology research institutions, and innovative drug development fields provided answers from different perspectives.
From the perspective of technological evolution, AI pharmaceuticals are still in their early stages and have become a common judgment among multiple guests.

Sun Weijie, founder and CEO of Shenshi Technology, believes that although computer-aided drug design (CADD) and AI assisted drug design (AIDD) have been developed for many years, the application of AI in the entire drug development system is still limited. "If we compare the development of AI to the industrial revolution brought by the steam engine, AI pharmaceuticals may still be in a very early stage today. ”

It is also believed that AI pharmaceuticals are still in their early stages, but the judgment dimensions of Zheng Shuangjia, a researcher at Lingang Laboratory and associate professor at Shanghai Jiao Tong University, are different.
If placed in the traditional pharmaceutical industry cycle, AI pharmaceuticals have made phased progress; But from the perspective of AI technology development, the industry is still in a very early stage. ”Zheng Shuangjia pointed out.

He pointed out that after ChatGPT, AI truly entered a stage of rapid development. In the future, with the emergence of AI Scientists with more universal and autonomous reasoning capabilities, the impact of AI on drug development may no longer be just about improving efficiency, but may redefine the research and development model.

In fact, not only in the technical stage, the reason why AI pharmaceuticals are still considered to be in the early stages is also reflected in the fact that there are still a large number of complex scientific problems that have not been solved.
At present, the application of AI at the protein level is still a focus, including enhancing the drug resistance, activity, and selectivity of molecules. But in the future, AI needs to further expand to the cellular level, and virtual cells, digital twins, and other directions may become new breakthroughs. Complex biological issues such as non pharmacological targets still require the combination of AI and experimental systems. ”Tengmai Pharmaceutical co-founder and CEO He Qi pointed out.

For this stage, Lai Caida, co-founder, chairman, and CEO of Jitai Technology, provided a more macroscopic judgment. In his view, AI pharmaceuticals are not like a football game with clear first and second halves, but more like a long-term marathon, "even if it's just a marathon, it's just starting

Lai Caida believes that the scale of investment in the global AI pharmaceutical industry is still limited, and the industry is still in the early stages of exploration. In the future, with more capital, technology, and industry players entering, the development space of AI pharmaceuticals is still enormous. The current global investment in AI pharmaceuticals may only be in the billions to 10 billion US dollars range, which is very small. With the continuous growth of investment and player numbers in the future, the explosive power of AI pharmaceuticals will far exceed current imagination. ”

But early on doesn't mean there's no competition. The performance of global leading companies also shows that AI pharmaceuticals are transitioning from technological exploration to industry validation. In the first half of this year, Schr ö dinger and Recursion both released underperforming annual reports, with Recursion having accumulated losses of over $2 billion since going public in 2021.

Regarding this, Ren Feng, CEO and CSO of Yingsi Intelligent, emphasized that from a technical perspective, AI pharmaceuticals are still in the early stages; But from the perspective of enterprise development, the industry has reached a critical juncture of life and death.

02. With the influx of big model giants, what are the barriers to AI pharmaceuticals?

The hot track has never been dominated by a single player.

As the AI pharmaceutical boom heats up, top players in the big model industry are also accelerating their entry into the game, constantly reshaping the competitive landscape of the track.

Since the beginning of this year, Anthropic has invested $400 million to acquire AI biotech startup Coefficient Bio and launched Claude Science for scientific research scenarios; In China, ByteDance split AI pharmaceutical business for independent financing, and Baitu United and Heping Pharmaceutical established AI pharmaceutical company. A series of actions indicate that large model enterprises are accelerating their expansion into the field of life sciences.

The entry of big model giants has also brought new thinking to the industry: Will the value of AI pharmaceutical companies be replaced by big model companies in the future? What are the barriers for AI pharmaceutical companies?
In Zheng Shuangjia's view, this concern is not valid.

He believes that the layout of AI for Science by big model companies is not simply about entering a specific industry, but more importantly, they hope to further enhance their basic model capabilities through complex tasks and high-value data discovered in science. Life sciences naturally require long chain reasoning, which is also an important scenario for testing the capabilities of large models.

In other words, big model companies focus on how to make their models have stronger scientific capabilities, while drug development itself still needs to focus on long-term accumulation around specific diseases, targets, and drug forms.

Of course, this also reveals that AI pharmaceutical companies do not need to compete with big model companies for the underlying models themselves. As Sun Weijie said, "For AI+biopharmaceutical companies, there is no need to consider the basic model, as no one can surpass the top general basic model manufacturers. ”

Focusing on drug research capabilities itself is still the way out for AI pharmaceutical companies. In Sun Weijie's view, what AI pharmaceutical companies really need to compete in the future is not the basic model itself, but how to utilize the capabilities of large models to establish an AI capability system for real R&D scenarios.

This system includes at least three levels: first, intelligent agents and research tools to help AI participate more effectively in scientific tasks; The second is the computational solving ability for life sciences, which can solve complex biological problems such as proteins and cells; The third is the long-term accumulated R&D experience, data resources, and know-how of the enterprise itself.

Ultimately, drug development still needs to form a closed loop in the physical world. Sun Weijie believes that AI can improve research and development efficiency, but what truly determines enterprise value is whether it can find good assets and transform AI capabilities into real drug results.

In fact, this is also the key difference between drug development and ordinary AI applications.

Ren Feng pointed out that there is no single answer to drug development, and model predictions still need to be verified through experiments, clinical studies, and real drug assets. Therefore, the core competitiveness of AI pharmaceutical companies is not model capability itself, but validated research and development capabilities, as well as the ability to integrate AI with drug development processes.

In addition to the research and development loop, long-term accumulated data and domain knowledge will also become important advantages for AI pharmaceutical companies.

Zhang Xiao, founder and CEO of Yuansi Peptide, pointed out that life sciences are highly complex, and different diseases, drug forms, and technological routes require long-term data accumulation and professional expertise.Especially the private domain data formed around specific diseases and drug forms, as well as the understanding of key scientific issues, cannot be quickly replicated by large model enterprises.

Therefore, AI pharmaceuticals will not move towards a simple substitution relationship. The innovative pharmaceutical industry itself is a highly diversified industry. Over the past few decades, different technological routes and business models have continuously nurtured new biotechnologies, and large pharmaceutical companies have also continuously obtained external innovation through cooperation and mergers and acquisitions. In the words of He Qi, the big model race is often a winner takes all race, but the AI pharmaceutical race is different, and there are still opportunities for latecomers to flourish.

In the future, large model enterprises may provide stronger foundational capabilities, while AI pharmaceutical companies will establish their own advantages around biological problems, data accumulation, and R&D loops. The two are not zero sum competition, but will work together to push AI deeper into the life sciences in the new industry chain.
 

03. How to layout AI pharmaceuticals?

After seeing the track pattern clearly, the problem facing AI pharmaceutical companies is also more realistic: as AI continues to improve research and development efficiency, how can companies formulate better pipeline development strategies and achieve commercial value transformation?

In the past, biotech often focused on a few core assets to advance research and development, from early detection all the way to clinical practice. However, AI is changing this mode - after the improvement of R&D efficiency, enterprises have the opportunity to lay out more pipelines at the same time, and improve the probability of R&D success through a richer portfolio.

But an increase in the number of pipelines does not necessarily mean a natural increase in value, and may even create new competitive pressures. How to avoid falling into homogeneous competition and find assets with truly differentiated value has become a question that enterprises need to consider in the next stage.

The key to future competition is not just generating molecules, but creating differentiated value. With the improvement of early R&D efficiency, simply providing molecular discovery capabilities may become increasingly difficult to form long-term advantages, and companies need to further extend to higher value links, "said Lai Caida.

Zhang Xiao also has a similar viewpoint. In his view, companies need to find truly valuable scientific problems and solve key biological challenges. As AI lowers the early R&D threshold, the deep accumulation formed around specific diseases, drug forms, and biological problems will become an important source of differentiated competition for enterprises.

In fact, current AI pharmaceutical companies are also exploring different development paths. A type of enterprise chooses to expand its asset portfolio and improve research and development efficiency by simultaneously promoting multiple pipelines; Another type of enterprise establishes vertical field advantages around specific disease areas or new forms of drugs.

Regardless of which path is chosen, innovative drug research and development is always a long-term, high investment project. How to utilize limited resources to promote asset development has become a problem that AI pharmaceutical companies must face.

Lai Caida believes that in the future, enterprises may not necessarily need to focus on a single asset for long-term promotion, but can allocate assets through various methods such as license out, Co co (joint development), NewCo, etc., to improve asset utilization efficiency.

Regarding the selection of specific development strategies, Ren Feng frankly stated that the core is not whether the assets are generated by AI, but depends on the company's own resources and strategic priorities. Taking Yingsi Intelligent as an example, the company's advantage lies in early drug discovery, so it is more inclined to leverage front-end research and development capabilities, while clinical development and commercialization are completed through partnerships.

In fact, as AI capabilities continue to integrate into the drug development process, the development model of AI pharmaceutical companies is also presenting more possibilities.

Zheng Shuangjia analyzed that the current model centered on pipeline assets will still be mainstream, but there may also be more platform based enterprises in the future. Companies such as Chai Discovery are exploring new models of serving pharmaceutical companies through AI tools and platform capabilities, with the value of helping R&D teams improve efficiency rather than directly participating in drug asset development.

Whether it's an AI platform, a drug pipeline, or a collaborative development model, it essentially depends on the company's own genes. Different enterprises need to choose their path based on their own advantages. "The most important thing is to choose the direction they are best at and achieve industry leadership," Sun Weijie concluded.
Perhaps in the future, AI pharmaceuticals may not form a single development paradigm. Some people choose to become AI driven biotech and promote commercialization around innovative assets; Some people focus on research and development platforms to provide basic capabilities for the industrial chain; Some people are also exploring new asset trading models.

The benefits brought by AI are not only the improvement of research and development efficiency, but also a re adjustment of the division of labor in the innovative pharmaceutical industry chain. Finding the right position may give you the opportunity to enter the next station and become a long-term player.

04. Summary of freehand brushwork

AI pharmaceuticals have no finale.

Technology will iterate, players will change, and business models will constantly be restructured. But no matter how the big model develops, the ultimate determinant of enterprise value is still whether it can solve key biological problems and transform AI capabilities into validated pharmaceutical assets.

From AI assisted research and development, to AI Scientists, and even stronger scientific intelligence in the future, AI is redefining the division of labor in the innovative pharmaceutical industry chain. Big model enterprises, AI pharmaceutical companies, traditional biotech and pharmaceutical companies will seek their own positions in the new industrial ecosystem.

For Chinese AI pharmaceutical companies, this is also a new exploration. Relying on the development of local innovative drug industry and the continuous accumulation of data and technical capabilities, Chinese enterprises are exploring different paths such as drug discovery, R&D platform and asset development to participate in global AI pharmaceutical competition.

In the next stage, the key to industry competition is no longer who has the largest model, but who can truly transform AI into the ability to create drug value.