(Yicai) Aug. 31 -- The use of artificial intelligence throughout the whole drug development process is entering the validation phase in China. As demand for AI talent in this field increases, competition intensifies, and salaries rise.

Under the traditional development model, advancing a drug from target discovery to a pre-clinical candidate takes an average of four and a half years. AI can shorten this timeline to under two years, as the leap in capabilities brought about by AI models has shown the pharmaceutical industry the true potential of AI-driven drug development.

“AI is applied primarily to boost early-stage research and development efficiency, specifically in pre-clinical molecular discovery,” said Ren Feng, co-chief executive officer and chief scientific officer of Insilico Medicine. “With the same budget, we can discover more candidate molecules, thereby doubling the number of drug candidates entering the pre-clinical stage.”

The application of AI is becoming a standard operational requirement for innovative pharmaceutical companies, Li Sichun, partner at Chinese healthcare headhunting firm Healthview, told Yicai.

Demand for professionals in AI-driven drug development keeps growing, Li said. Demand is highest for professionals in computational drug discovery and engineering applications, especially those skilled in drug design, structural optimization, and virtual experimentation.

“A few years ago, drugmakers mainly poached talent from internet firms who understood algorithms but lacked pharmaceutical expertise, and then had them undergo a period of mutual learning and integration with scientists,” Li explained. “Over the past two years, these pioneer AI talents specialized in drug development have become a prime target for headhunters, as they often receive two to three offers immediately after they quit a job.”

“It remains very difficult to find interdisciplinary talent who understand both AI and drug R&D,” Ren said. “The vast majority of our workforce has been cultivated in-house, which involves hiring AI specialists and biomedical professionals and training them through hands-on project experience for one to three years to forge them into interdisciplinary experts.”

This interdisciplinary talent pool remains far too small to meet the industry’s surging needs, according to Li. “Market demand for AI-driven drug development talent is twice the size of the existing talent pool, and the number of job vacancies is four to five times the number of professionals willing to change jobs.”

Qualified candidates have been extremely hard to find since last year, and competitors frequently snatch them up before offers can be finalized, Li noted. For instance, Li said that the three candidates under consideration for an AI-driven drug development role were all poached by competitors, as he is still searching for a candidate for this position after 10 months.

Healthview advised the employer to raise the salary for this position by 30 percent, warning that otherwise, the next promising candidate would likely be snatched away just as easily.

Annual salaries offered by some drugmakers for AI talent hover around CNY500,000 (USD74,330) for managers and reach CNY1 million to CNY1.5 million (USD148,660 to USD222,990) for directors. Project leaders can ask for more than CNY2 million.

Due to the tight talent supply, recruitment cycles for AI-driven drug development positions are lengthening, averaging two to three months. Filling higher-paid project leader roles can take six to 12 months.

Insilico has encountered numerous collaborative opportunities for new drug development this year but had to turn down some due to a lack of manpower, Ren admitted to Yicai. The company plans to hire more talent while implementing measures to retain its current workforce.

AI-driven drug development professionals with three to four years of experience are exceedingly scarce on the market, so Insilico’s existing employees are at constant risk of being poached by competitors offering premium salaries, Ren noted, adding that this requires the company to roll out various incentive programs to retain them.

Its early investment in AI-driven drug development has allowed Insilico to reap a windfall of orders. Since the beginning of this year, it has signed licensing deals over its innovative drug candidates worth about USD7.3 billion with multinational pharmaceutical giants, including Servier, Eli Lilly, and Takeda Pharmaceutical.

Insilico’s revenue surged 287 percent to USD106 million in the first half from a year earlier, according to its semiannual earnings report. It turned a net loss of USD19.2 million into a net profit of USD35.5 million in the period.

Ren believes that AI-driven drug development has not yet reached its ‘DeepSeek moment,’ as the pharmaceutical sector will only truly break into the mainstream when AI-developed drugs win regulatory approval and hit the market.

“There is no data to support whether AI can drastically improve the success rate of new drug R&D, as no drug developed entirely with AI has yet hit the market,” Ren said.

“I firmly believe that the success rate of AI-driven drug development will outstrip traditional empirical models,” he noted. “Traditional R&D relies heavily on human experience, whereas AI-driven drug development adds a data-driven process to that.”

Editor: Futura Costaglione