Life sciences face ‘scientific content crisis’ in AI adoption

The Pistoia Alliance, a global not-for-profit promoting collaboration in life sciences research, has released data highlighting a growing ‘scientific content crisis’ that is limiting the accuracy and adoption of artificial intelligence (AI) in R&D.

The findings, gathered from a poll conducted at the Alliance’s annual US conference in Boston, showed that 27 percent of four life science professionals do not know what scientific content their organisation’s AI or large language model (LLM) systems use – or rely solely on titles and abstracts. Meanwhile, only around 36 percent of respondents said they were incorporating internal documents into AI models.

This lack of comprehensive, traceable evidence is undermining confidence in AI outputs, as many systems are being built on incomplete datasets. The conference collected together more than 170 experts from pharma, technology and academia to explore AI challenges and the governance measures needed for safe implementation.

Data and governance gaps raise concerns

“It’s clear from discussions at the conference that many AI models are not yet drawing on the full range of scientific evidence needed to deliver authoritative results. Many organisations are still in a learning phase when it comes to both data and governance and, given the stakes for patient safety, that cannot be ignored,” said Senior Director Neal Dunkinson of the Copyright Clearance Center.

38 percent of respondents said their copyright and licensing policies are unclear or not even enforced.

Dunkinson highlighted further risks, noting that 38 percent of respondents said their copyright and licensing policies are unclear or not even enforced. “This means many could also be at risk of fines in an already costly drug development process. To ensure models are grounded in the highest-quality and most complete scientific datasets, the industry must ensure any datasets being used are AI-ready: meaning properly structured, licensed and transparent,” he said.

The need for benchmarking and standards

The poll also emphasised the importance of stronger benchmarking and governance for AI agents. half of respondents identified the absence of shared verification standards as the biggest barrier to agent adoption.

In response, Robert Gill, Agentic AI programme lead at the Pistoia Alliance, encouraged conference attendees to become founding members of the Alliance’s agentic AI project. The initiative aims to develop standards for safe, scalable AI and ensure organisations have full visibility over the data their AI models are learning from.

Conference highlights: AI in practice

Several sessions demonstrated practical applications of AI in life sciences:

• Accelerating clinical trials and R&D: EPAM presented AI-driven streamlining of clinical operations, the Michael J. Fox Foundation illustrated how knowledge graphs can speed Parkinson’s research and AbbVie discussed AI’s role in improving pharmacovigilance.
• Roundtable on AI adoption: Elsevier convened experts from companies including Eli Lilly, Pfizer, Bayer, J&J and Takeda to discuss implementing AI tools in real-world research settings. The consensus was that AI adoption depends on problem-led, intuitive design and seamless workflow integration.
• Change management and skills shortages: Representatives from Eli Lilly, Kalleid, Elsevier, Takeda and Ziffo emphasised that AI success relies on people and incentives as much as technology. This echoed findings from Pistoia’s Lab of the Future survey, where 34 percent cited a shortage of skilled talent as a barrier to AI adoption.

Global concerns and next steps

“It’s notable that the same concerns around AI trust, transparency and skills were raised at both our US and European conferences. These issues are clearly universal across the life sciences community,” said Dr Becky Upton, President of the Pistoia Alliance.

She added that collaboration on common standards, data quality and practical implementation is key to advancing the industry with confidence. “The Pistoia Alliance exists to facilitate this collaboration, and we’re excited to carry these discussions into our spring meeting in London.”