Major pharmaceutical companies are reporting that artificial intelligence is significantly accelerating early-stage drug discovery, with some saying the technology has reduced the time required to design and develop potential medicines by as much as half.
Companies including GSK, Bristol Myers Squibb, Novartis, Eli Lilly, Novo Nordisk, AstraZeneca and Sanofi are using AI to design molecules, identify promising drug targets and assess potential safety risks.
The growing adoption of the technology is also drawing attention to Anthropic, the company behind the Claude AI models, which is expanding its involvement in pharmaceutical research through partnerships, investments and its own drug discovery initiatives.
However, scientists say AI is speeding up parts of the process rather than eliminating the lengthy clinical trials and regulatory checks required before medicines can reach patients.
Pharmaceutical companies report faster drug discovery
Drugmakers say AI is moving beyond administrative tasks, such as preparing regulatory documents, and is increasingly being used in core scientific research.
Sanofi said an AI-supported approach helped it save at least a year in developing its experimental asthma treatment, lunsekimig, by supporting early biological research and clinical decisions that allowed the company to skip a dose-selection study.
Novartis said AI tools helped researchers identify five potential targets for kidney disease treatments. The company also reported reducing the average time required to move from the start of a research project to a drug candidate from about three years to between 18 months and two years.
Novo Nordisk said AI had reduced the number of design, manufacturing and testing cycles for new molecules from approximately eight to three. It also cut the duration of each cycle from eight weeks to four.
Bristol Myers Squibb, meanwhile, said one of its experimental sickle cell disease treatments would not have been developed without its AI systems.

Anthropic emerges as a potential pharmaceutical rival and partner
Anthropic is attracting interest from drugmakers as it expands the use of its Claude models in scientific research.
The company has established partnerships with pharmaceutical and biotechnology businesses, including Bristol Myers Squibb, and has introduced Claude Science, software designed to support scientific research.
Novo Nordisk has also partnered with Anthropic to explore how its AI models can support research and development and accelerate the discovery of new medicines.
The company is positioning itself as both a potential collaborator and competitor to established pharmaceutical businesses. Some drugmakers see AI companies as technology providers and research partners, while others are increasingly concerned about their potential to compete directly in drug development.
Anthropic has also expanded its own drug research activities, including work on treatments for rare diseases.
AI cannot yet guarantee faster drug approvals
Despite the reported progress, scientists caution that faster discovery does not necessarily mean medicines will reach patients sooner.
Drug development traditionally takes between 10 and 15 years and can cost billions of dollars. Nearly nine out of 10 drugs developed through conventional methods fail during clinical trials, according to industry figures cited by Reuters.
Researchers also say the number of AI-supported drug candidates entering clinical testing remains relatively small. A March assessment identified 129 candidates across 69 companies, compared with approximately 117 candidates eight months earlier.
Experts warn that clinical trials, which test the safety and effectiveness of treatments in people, and regulatory approval processes will continue to limit how quickly new medicines can reach the market.
It also remains uncertain whether drugs developed with AI will ultimately have higher success rates than those discovered through traditional methods.

Pharmaceutical firms remain cautious about sharing data
Access to proprietary research data is emerging as a key issue in the relationship between pharmaceutical companies and AI developers.
Drugmakers possess decades of scientific research, clinical information and experimental results that can help train and improve AI systems. However, some companies are reluctant to share sensitive internal data with external technology providers.
Novo Nordisk and Sanofi have said they do not provide proprietary data to major AI companies for training their models.
Sanofi’s global head of computational and AI strategy, Matt Truppo, said AI systems could become less effective without access to relevant data.
The company expects AI to reduce early drug discovery timelines by 20 to 25 per cent within two years, potentially reaching 50 per cent within five years across more than 100 drug development programmes.
For pharmaceutical companies, the challenge will be determining how to benefit from increasingly powerful AI tools while protecting valuable research data and maintaining control over drug development.
Discover more from LN247
Subscribe to get the latest posts sent to your email.

