Why Younger Clinical Researchers Are More Wary Of AI Than Their Elders
By John Oncea, Chief Editor, Clinical Tech Leader
Every previous wave of workplace technology has followed the same adoption curve: younger staff pick it up fast, older staff resist. Liam Devenish, a physician and former clinical trial sub-investigator now researching AI strategy and health policy at the London School of Economics, has noticed AI breaking that pattern, and he thinks the reasons matter for anyone planning an AI rollout in a research organization.
An Adoption Curve Running Backwards
“I definitely see that trend,” Devenish said when asked whether he’d observed younger researchers more resistant to AI than their older colleagues. “Yeah, absolutely. I think it’s true.” For an industry used to assuming generational tech-fluency maps neatly onto adoption willingness, that reversal is worth sitting with before rolling out an AI tool and assuming the youngest staff on the team will be the easiest adopters.
Three Reasons, Not One
Devenish offers a layered explanation rather than a single cause. The first is economic and political: he sees AI, for a meaningful share of younger researchers, “as an outflow of a hypercapitalist meritocracy that has concentrated wealth and privilege amongst a very small number of people and is being deployed to concentrate further wealth” – resistance rooted in who benefits from the technology, not how well it works.
The second is a values shift. “Generally younger people care more about the idea of inclusion, at least in this generation, and see the importance of that,” he said. “And I think AI is a real threat to that.” He connects this to a broader concern about creative and intellectual displacement, a sense that AI threatens “our being able to create in the world, whatever create means for you.”
The third is about trust in information itself. Younger researchers came of age amid what Devenish calls “a rapid dissemination of information” and a rising “barrier to understanding what truth is” – a landscape AI adds another layer of uncertainty to. Older generations, he suggests, built their information habits around “journalistic integrity” and slower-moving news cycles, which left them “less primed to that risk.” A new technology like AI, in that framing, doesn’t trigger the same alarm for someone who didn’t grow up needing to interrogate every source.
What This Means Beyond The Interview
Devenish doesn’t frame this as a reason to slow AI adoption in clinical research, and his companion framework for deploying AI responsibly – linked to at the end of this article – makes clear he sees real, narrow use cases worth pursuing now. But the pattern he describes has a direct operational implication that’s easy to miss: an organization can build a well-scoped, well-sandboxed AI tool that clears every technical and value bar and still meet resistance from the staff it depends on to make the rollout work for reasons that have nothing to do with the tool’s performance.
For sponsors and CROs planning AI deployments, that suggests change management needs to do more than train people on a new interface. If the resistance is rooted in concerns about equity, information trust, or who the technology ultimately serves, no amount of workflow documentation addresses it. It’s a conversation about purpose and governance, not a training module, and Devenish’s read suggests research organizations that skip that conversation may find their newest, most digitally fluent staff are the hardest group to bring along, not the easiest.
This is a companion piece to A Framework For Deploying AI In Clinical Trials Without The Hype.