Spotlight Brief
Co-Skilling as an Absorption Strategy

Translating Zhang et al. (2025) for board and investor audiences in health and life sciences.

 

This brief reads one recent study through a board and investor lens. The study is about fear. Specifically, it's about what keeps people from spiraling into fear about their jobs when AI lands in their workplace, and what an organization can actually do about it.

​

What the study found

Zhang and colleagues (2025) surveyed 437 employees across manufacturing, healthcare, technology, banking, and retail firms, and tested a model built on social learning theory and the job demands-resources framework. Their answer centers on something they call Co-Skilling, learning to work with AI as a shared, collaborative activity rather than a solitary scramble.

The study traces how Co-Skilling lowers job insecurity through three buffers. The first is perceived organizational support, the employee's read on whether the company has their back. The second is mental wellbeing. The third is skill confidence, the quiet belief that you can handle what the technology throws at you. Of those, organizational support did the heaviest lifting. The headline for a leader is short. AI capability is something you buy. The capacity to absorb it is something you build, and Co-Skilling is one of the few build mechanisms with real evidence behind it.

​

Why it lands harder in healthcare

In most sectors a frightened or disengaged employee is a productivity problem. In health and life sciences it becomes a safety problem. A nurse who doesn't trust the triage model, a claims adjuster who has quietly checked out, a clinician who works around a new decision-support tool. Each of those is a clinical and financial exposure, not a morale footnote.

There's a second reason this matters at board level. Clinical judgment is deeply tacit knowledge. It lives in experience, not in manuals, and it doesn't transfer through a memo. When AI arrives and people fear for their roles, the first asset to walk out the door is exactly that tacit expertise. Powell and Snellman (2004) made the point two decades ago that the gains from a new general-purpose technology show up only when the organization adapts around it. Zhang and colleagues (2025) give us the human mechanism for that adaptation.

​

What boards should do

Boards don't implement programs. They set expectations and they watch the right numbers. So the translation is about what to expect and what to watch.

Sequence matters first. The study found through a necessary condition analysis that an organization has to clear a floor of support before broad skill-building does much good. You cannot reskill your way out of a workforce that doesn't believe the organization has its back. Build the support floor, then scale the learning on top of it.

Readiness is never uniform, so the program cannot be uniform either. A single mandatory training pushed at everyone wastes money on the ready and frightens the unready. Segment it. Give the hesitant group fundamentals and low-stakes practice, and give the confident group room to lead pilots and mentor peers.

Wellbeing belongs in the design rather than bolted on afterward. In a sector already carrying heavy burnout from documentation and regulatory weight (Sinsky et al., 2016), that finding is structural rather than soft. The thread running through all of it is the relocation thesis. AI migrates the status of work from execution toward judgment and oversight, an arc that mirrors the augmentation pattern Autor (2015) traced across earlier waves of automation. Co-Skilling is how you make that migration survivable for the people living through it.

​

The diligence lens

For our investor clients we add one more move. When assessing a target's AI story, don't stop at the technology stack. Ask what the company has built on the Receptance side. A target with a strong Co-Skilling culture is carrying an intangible asset that doesn't show up on the balance sheet and that compounds over time. A target that has lit up its model roadmap while hollowing out its workforce is carrying a hidden liability. That's why we treat Receptance as a board-level concern rather than an HR footnote.