Culture and AI Adoption: Why Disengaged Teams Sink Healthcare AI

The underwhelming reception of a healthcare AI rollout typically centers around the subpar workings of the tool itself. The tool was clunky. The integration was rough. Clinicians worked around it. All of these factors are important to understanding a poorly implemented AI rollout; however, they are not the whole story. All of these issues sit downstream of whether the potential to adopt such a tool remained for the intended users.

This was the main concept of a conversation that took place in the Signal Room with Carly Caminiti, a leadership coach. Caminiti noted that the exhaustion of a company's staff can reframe perceptions of AI tool implementation. For exhausted teams, AI tools may feel like an administrative burden. Consider the case of an exhausted nurse that is close to leaving: an administrative tool that is meant to aid the nurse, and other staff, in managing their interactions and work can have adverse effects and can worsen their situation. It is common at Hutchins Data Strategy Consultants to encounter this situation, where an otherwise capable system fails to explain why the tool is not adopted and integrated.

Leadership's Blind Spot

Leadership often misunderstands what a 'stable, well-staffed workforce' means when compared to the actual experience on the front lines. As Caminiti describes it, leadership looks at the staffing numbers and concludes the unit is fully staffed. Conversely, the nurse assigned to this unit hears this statement and often reacts with disbelief. Many of the healthcare workers that Caminiti had spoken to had never once been told that their nursing floor was 'fully staffed.'

Software designed to expose this gap often conceals it instead. Most employees may describe their workplace with a positive overall rating, while among the employees themselves there is a division, one cluster giving a high rating and another cluster giving a low rating, with an overall rating that conceals this division and appears positive. She also cited an earlier Gallup statistic, which stated that around 20% of employees describe themselves as engaged, while the remaining 80% are somewhere along the spectrum of disengagement. For leadership to gain any meaningful insight, they must forgo the conventional route of reading the numbers and instead explore what the disengaged employees are experiencing.

Perhaps the biggest challenge is that surveys have been designed with an inherent trust of the employees that is no longer warranted. When employees are unable to voice their truth, leaders ultimately miss the opportunities to adapt, pivot, and better serve their employees.

When a Staffing Problem Becomes a Design Problem

Caminiti placed boundaries on the departure of a single employee as opposed to a systemic issue. When multiple employees from the same team, or in the same position, leave, it is no longer a single employee problem. This is a leadership and design problem.

This has a lot to do with how employees ascend to the level of leadership. A former employee is the best in their field, be it in health care or as a tradesperson, and are subsequently promoted. However, being the best in one's field is not being the best in leading and managing the employees doing the work, and this promotion does not come with the management training to fill this gap. The quiet spiral she described happens when the newly promoted person suspects they did not deserve the role and ultimately leaves, due to a gap that the organization created, but has not named.

The gap has a cost that many organizations do not account for on their balance sheets. Caminiti relayed what senior leaders have told her directly — in some circles, firing barely registers as a decision, just a question of how quickly a person can be replaced. The replacement is costly: recruiting, onboarding, and the loss of institutional knowledge that has to be rebuilt. A decision that looks like a responsible financial decision is actually an accounting exercise that does not account for the total cost.

Burnout is Contagious: It Affects Patients Too

In her theory, burnout is something that develops over a significant period of time. It is the result of long-term chronic stress that goes unacknowledged by both the employee and the organization. It is also not self-contained. A few people in a team are almost about to leave, and can shift the entire team's dynamics.

Burnout also spreads to patients in the healthcare system. When employees do not feel supported and are overburdened and stressed out because of the high demand while working in an understaffed unit, the patients absorb their energy. They feel more rushed and strained, and they reflect that in their care. The dissatisfaction of the staff bleeds to the patients, and because of that, culture is directly linked to the work done in the organization.

Her discomfort with cutting budget for leadership training stems from her understanding of the damaging effects of a reduction in workforce. For the employees who remain after a layoff, the loss impacts their workplace safety and their psychological safety is eroded because they leave work wondering if they will be next. The organization is more likely to cut the budget at a time when they should be investing in the remaining employees. A manager who cannot have an uncomfortable conversation, and a team with a leadership crisis are absent from budget items.

Trust is Built Small, and AI Cannot Build It

If culture is a prerequisite for adoption, trust is a prerequisite for culture, and Caminiti offered some specifics on the creation of trust. Trust is not built through one singular event. It is cultivated through the consistency of small acts over long periods of time, and it resides in relationships with direct reports. If that relationship is neglected and communication stops, it is the first signal of trust diminishing.

Small signals cut both ways. If a leader sends a work-related text to a subordinate when the leader is not working, the official communication policy is not necessarily negated, but the employee is still compelled to respond, often due to the unspoken boundaries of the relationship. Caminiti believed that most senior leaders fail to notice the disconnect. They assume that modeling good behavior is enough, without seeing that someone lower in the hierarchy may feel unable to do the same things their boss can.

This is when her perspective of AI becomes sharp, and where it aligns with adoption. Rather than viewing AI as the enemy, she believes it will push people back toward human connection — picking up the phone, sitting in a room together, and noticing what a screen cannot. What technology cannot do is listen to emotion, read a room, or empathize, and those are precisely the capacities that repair a team. An organization that hopes a new tool will paper over a trust deficit has the sequence backwards. The human work comes first, and the technology lands on whatever foundation that work has, or has not, built.

How Hutchins Approaches Culture and AI Adoption

We treat readiness as a question about people as much as systems. Before a model reaches a clinician's workflow, we look at whether the team has the capacity, the trust, and the shared understanding to absorb it — because a rollout into a depleted, disengaged unit confirms every reason to distrust AI rather than earning a reason to adopt it. That assessment connects to the AI literacy that lets staff judge a tool instead of fearing it, and to the operational alignment that keeps a data and AI strategy tethered to how the work is actually done. These themes run throughout The Signal Room podcast, where leaders describe what it takes for technology to land in a real organization rather than an idealized one.

Authoritative sources

• WHO — Ethics & governance of AI for health
• Agency for Healthcare Research and Quality
• HIMSS
• NIST AI Risk Management Framework
• JAMA Network