Artificial intelligence has rapidly become one of the most transformative technologies in the pharmaceutical industry. While it is far from replacing the expertise of scientists, AI is already accelerating the early stages of drug discovery and development by streamlining time-consuming research processes and reducing research and development costs.
According to a TD Cowen survey of 80 biopharma executives and industry experts, AI could reduce the time and cost of preclinical drug development by as much as 70%. The preclinical phase includes designing drug candidates, evaluating their biological properties, and conducting initial laboratory studies before human clinical trials begin.
Analysts believe this shift is driving demand for specialized software, genomic sequencing technologies, and advanced computational models, tools that could significantly increase the number of experimental drug candidates entering development over the next five years.
During the discovery phase, AI analyzes massive biological datasets, identifies promising drug targets, and predicts how different compounds are likely to perform. This enables researchers to evaluate far more therapeutic candidates in a shorter period of time. Nevertheless, laboratory experiments remain essential for validating the safety and efficacy of these compounds, and no AI-generated prediction can replace experimental evidence.
One of AI’s most significant contributions is the expansion of in silico platforms, computer-based systems capable of performing thousands of virtual experiments within seconds. These platforms can predict properties such as toxicity, stability, drug metabolism, and even potential drug-drug interactions before laboratory testing begins.
The survey also found that demand for software capable of simulating human biological processes is expected to experience the fastest growth through 2028. Beyond predicting drug interactions, these tools could help optimize dosing strategies for vulnerable populations, including newborns and pregnant women.
According to the report, the number of new drug development programs could increase by more than 10% over the next three to five years. At the same time, increased investment in laboratory infrastructure and AI-powered technologies is expected to add roughly $1 billion in costs across the pharmaceutical industry.
These technological advances are also being reinforced by new U.S. policies aimed at reducing the use of animals in biomedical research. The shift is expected to accelerate adoption of computational models, three-dimensional human tissue models, and other alternative technologies capable of evaluating the safety and toxicity of drug candidates before they enter human clinical trials.
Despite this progress, experts caution that AI has not yet guided a drug from initial design through final approval by the U.S. Food and Drug Administration (FDA). As a result, questions remain about how much the technology will ultimately improve patient outcomes.
A major concern is that current AI models still cannot fully capture the complexity of human biology and the wide variability in individual patient responses. Consequently, some researchers believe the failure rate in clinical trials—which is currently estimated at around 90%, may remain high despite advances in AI.
Meanwhile, global competition in AI-driven drug discovery is intensifying. China’s rapidly expanding biotechnology sector, supported by lower development costs and faster research timelines, has attracted substantial investment and emerged as a major competitive challenge for U.S. pharmaceutical companies.
Overall, experts agree that AI is unlikely to replace scientists, but it has the potential to fundamentally reshape how medicines are discovered and developed. By shortening research timelines, lowering development costs, expanding the number of compounds that can be evaluated, and improving decision-making, AI is poised to transform pharmaceutical R&D. Ultimately, however, its success will continue to depend on clinical trial outcomes and the real-world performance of new medicines in patients.
Re: ax