Drugmakers turn to AI to speed trials, regulatory submissions
Artificial intelligence has yet to crack the toughest challenge in drug development—discovering breakthrough molecules—but it is already accelerating the routine, time-consuming work that surrounds clinical research and regulatory submissions. From finding trial sites and participants to drafting documents for regulators, leading pharmaceutical and biotech companies say AI is helping shave weeks off labor-intensive workflows and streamline paths to market.
From ambition to execution
Bringing a new medicine to market can take a decade and cost billions. Many drugmakers are betting that AI can lift success rates and compress timelines, complementing traditional discovery efforts. Partnerships and tool investments are proliferating across the sector, with some companies aligning closely with computing leaders to unlock the technology’s potential.
Agentic AI—systems that can operate with minimal human oversight—could boost clinical development productivity by roughly 35% to 45% over the next five years, according to industry estimates. While the ultimate goal is better, faster drug discovery, companies are deploying AI across the development lifecycle to make measurable progress now.
Turbocharging the “messy middle” of development
Executives from global players and emerging biotechs alike describe using AI to manage the complex web of documentation required by regulators. Clinical, safety, and manufacturing records must be compiled, cross-checked, and kept consistent across regions—work that has often required expensive external support. Automating document tracking and harmonization can reduce costs and shorten review cycles.
Investors are also targeting the operational bottlenecks that stall promising programs. Trial enrollment is often a “leaky funnel,” with potential participants dropping out at multiple stages. Startups are applying AI to patient outreach, education, screening, and scheduling to keep trials on track and improve diversity and retention.
Adoption is rising, impact is emerging
Across the industry, AI tools—especially large language models—are now commonly used for administrative tasks such as drafting, summarizing, and data extraction. Measuring bottom-line impact will take time, however. Analysts suggest it may be one to three years before investors can clearly quantify how much AI has accelerated development timelines, since outcomes vary based on how and where tools are deployed.
Case studies: fewer weeks, better choices
Novartis leaned on AI in a 14,000-person, late-stage cardiovascular outcomes study for its cholesterol therapy Leqvio. A site-selection process that typically takes four to six weeks was condensed into a two-hour meeting by using AI to identify higher-performing sites. Novartis completed enrollment with only 13 participants above target, highlighting the precision of the approach. Company leaders say time savings like these can compound into months gained across an entire program, reframing AI as “augmenting intelligence.”
GSK has combined digital tools and AI to cut manual data collection and aggregation and accelerate enrollment, with an aim to speed all clinical trials by 15%. The company estimates it saved around £8 million in late-stage studies of the asthma drug Exdensur, which recently secured U.S. approval.
Automating the post-trial grind
Danish antibody developer Genmab plans to deploy agentic AI powered by advanced chatbots to support key clinical development priorities. The goal is to automate post-trial tasks—analyzing datasets, converting results into graphs, tables, and figures, and assembling clinical study reports—to reduce cycle times and free experts for higher-value work.
German radiopharmaceuticals firm ITM has built AI tools that can convert long trial reports into standardized U.S. FDA templates, a process that traditionally takes weeks and multiple staff. While not yet deployed at scale, the approach could materially reduce manual formatting and review.
The near-term promise and the long-term prize
R&D leaders report that AI is already delivering productivity gains in trial operations and regulatory preparation. The enduring question is when AI will produce entirely new, high-impact medicines. Some executives believe those first “AI-designed” molecules are already moving through pipelines.
For now, the practical wins are accumulating: smarter site selection, faster enrollment, lighter documentation burdens, and quicker turnarounds on regulatory-ready materials. If current momentum holds, AI may not just speed the journey from lab bench to bedside—it could reshape how the industry runs clinical development end to end.
