SAN FRANCISCO, January 26, 2026, 05:15 (PST)
- Drugmakers report that AI is reducing the hours needed for trial logistics and handling regulator paperwork
- Novartis reported that AI cut the trial site selection process for a 14,000-person Leqvio study down to just a two-hour meeting
- Brendan Smith from TD Cowen noted it might be 1–3 years before investors really understand the impact
Drugmakers are increasingly relying on artificial intelligence to accelerate clinical trials and regulatory filings, slashing weeks off processes that often slow down development, according to executives and analysts. While AI hasn’t yet cracked the code for discovering groundbreaking new drugs, it’s already making a dent in the “messy middle” of drug development. 1
This matters since the timeline remains brutal. Companies report that developing a drug can take around a decade and cost close to $2 billion. Even when results seem promising, the slow grind of trial setup and paperwork can drag timelines out much longer.
AI deals are pouring in, like Eli Lilly teaming up with chipmaker Nvidia, as drugmakers explore tools to boost success rates and cut timelines. McKinsey reported last year that “agentic AI”—systems capable of handling tasks with minimal human input—might increase clinical development productivity by 35% to 45% within five years.
Teva CEO Richard Francis called the early wins deliberately boring. He emphasized that all other aspects of drug development must be “as efficient and as small as possible,” highlighting digitization and process improvements as key drivers that “make a difference.”
Paperwork remains a huge challenge. Companies like AstraZeneca report managing thousands of pages for regulators, covering clinical, safety, and manufacturing documents, all while ensuring consistency across different regions.
AstraZeneca Chief Financial Officer Aradhana Sarin said teams frequently need to compile, cross-check, and harmonize those documents across different regions, sometimes relying on external contractors to get the job done.
Some of the funds are targeting what investors refer to as bottlenecks rather than breakthroughs. Andreessen Horowitz general partner Jorge Conde mentioned he’s supporting solutions for drug development’s “messy middle,” including a $4.3 million investment in the startup Alleviate Health.
Conde described trial enrollment as a “leaky funnel,” where potential participants drop off at various stages. He said Alleviate is leveraging AI to bolster outreach, education, screening, and scheduling efforts.
TD Cowen analyst Brendan Smith noted that large language models—chat-based AI like Microsoft’s Copilot—are increasingly used for administrative work in pharma. However, he added it could be one to three years before investors can accurately gauge their impact on speeding up drug development. The amount of cost savings will depend heavily on how each company uses these tools.
Novartis gave a specific example. Chief Medical Officer Shreeram Aradhye explained that in 2023, the company deployed AI during the rollout of a 14,000-participant late-stage outcomes trial for their cholesterol drug Leqvio — with “late-stage” referring to one of the last major patient tests before seeking approval.
Aradhye explained that AI shrank the usual four-to-six-week trial site selection process—choosing hospitals and clinics for the study—down to just a two-hour meeting. It also helped Novartis close enrollment with only 13 patients over its target. “AI becomes augmenting intelligence, not artificial intelligence,” he noted.
GSK revealed it’s leveraging a combination of digital and AI technologies to cut down on manual data gathering and boost enrollment, targeting a 15% acceleration in clinical trials. According to a spokesperson, this approach saved roughly £8 million ($10.87 million) in late-stage asthma drug Exdensur trials last year. Notably, Exdensur received U.S. approval just last month. 2
Genmab announced plans to use Anthropic’s Claude chatbot-driven agentic AI to aid clinical development. The system will handle tasks like automating post-trial processes—turning data into graphs, tables, and clinical study reports, said Hisham Hamadeh, the company’s head of AI. Meanwhile, German radiopharmaceutical company ITM revealed it has figured out how to leverage AI to transform lengthy trial reports into FDA template formats, a move that could cut weeks off the timeline. However, ITM has not yet rolled out the technology.
Amgen research chief Jay Bradner revealed that AI is already speeding up drug development and the preparation of regulatory documents, even though the industry is still waiting for clear proof it can produce new medicines from start to finish. “What everybody’s waiting for is the AI drug,” he said. “I actually think those molecules are in pipelines right now.” 3
The argument for AI as a genuine time saver is still unfolding. Smith noted it might be years before investors can judge if quicker site selection, faster enrollment, and cleaner submissions really deliver results across a program. Some companies also admit their tools aren’t fully deployed yet.
