Empathetic Leadership Can Make or Break AI Adoption

At work, the human case for empathy is straightforward: Social connection is one of our greatest sources of well-being, and when workers feel it they grow happier and healthier. Research has now made the business case for it equally clear: Employees of empathic companies do better, working harder, collaborating more efficiently, and generating stronger ideas. AI, however, is putting human connection at risk. Many workers, deeply concerned about how the technology will change their jobs, are experiencing rising levels of mistrust and “FOBO,” or fear of becoming obsolete. A 2025 study found that when companies adopted AI, employees’ depression levels tended to increase over time. All of this highlights the importance of human connection in keeping employees healthy in the workplace as new technologies enter. But AI is also remaking the business case for empathy in real time. A growing body of evidence reveals that successful AI adoption depends on empathy, and companies who focus on human connection also lead the way in technical innovation. Many leaders are missing this point. Businessolver’s 2025 State of Workplace Empathy report found that 59% of CEOs believe empathy is non-essential (only “nice to have”) at work—and that’s up 12 points from 2024. Forty-nine percent said they don’t have time to connect with their employees, up 16 points from 2024. In their race to embrace AI, it seems, leaders are letting empathy go. In this article, drawing on new studies and my own research on empathy, I’ll explain why this is a mistake and how leaders can do better. When Empathy Goes Without empathetic leadership, it’s impossible to maximize AI’s potential. There are several reasons for that. Unempathetic leadership creates an AI-adoption gap. In a recent survey, CEOs expressed enormous confidence in AI. Eighty-one percent reported that their company had a clear AI policy, and 40% believed the technology was already saving their workers more than eight hours a week. This was breathtakingly out of touch with their own employees, two-thirds of whom said they saved two hours or less through AI, and only 28% of whom agreed that their company had clear strategies for using it. Leaders are making rosy, inaccurate predictions about what their employees are doing with AI in part because they don’t understand how people feel about it. A BCG survey last year found that 76% of executives believed their employees were enthusiastic about AI adoption in their organization; the real figure was only 31%. Senior executives stand to gain from new technologies and are empowered in strategizing about its use. Many employees, by contrast, feel unclear about the future and threatened by FOBO. After all, why would anyone feel enthusiastic about training their replacement? Leaders who don’t work to understand their people’s experience will remain unaware of these problems and unable to address them. Without empathy, workers innovate less. When people are cared for they feel safe, which gives them the confidence to take risks. This gives leaders enormous impact. A 2021 Catalyst survey found that 61% of employees who had empathic managers reported innovating at work, compared to just 13% of employees with unempathetic managers. Research reveals that empathic leaders decrease their team’s fear of uncertainty, allowing them to embrace, explore, and deploy new ideas. This is true of AI as well. Data from more than 2,000 employees found that people who felt safe were more likely to embrace AI tools, and in an MIT survey of 500 leaders, 84% observed a connection between psychological safety and tangible AI outcomes. Despite this evidence, many leaders have done little to encourage safety around new technologies, and most people remain wary. Most Americans believe AI will do more harm than good in their lives, and only 21% believe they can trust it most or all of the time. It’s impossible to expect companies to become AI-mature unless this changes. Unempathetic AI rollouts make the technology less efficient. In environments that lack care and safety, people try frantically to look busy, often producing lower quality work. AI has resurfaced this issue in new ways. Told to use it and expected to simply “be more productive,” employees have produced landfills of workslop: seemingly sensible AI output that lacks depth or value. Workslop is created in seconds but costs colleagues hours as they try to decipher it, and it damages collaboration in the process. This is just one symptom of a deeper risk. When AI is rolled out in the absence of empathy or trust, many employees actively oppose it. An enterprise-AI adoption survey conducted by Writer found that nearly a third of employees (and 44% of Gen Z workers) admit to sabotaging their company’s AI strategies—for instance, by feeding sensitive information to unauthorized models or tampering with outputs to make AI seem less effective. This is unethical but understandable: When placed in a zero-sum environment with a new technology, people will naturally protect themselves. The Empathy Advantage As this all makes clear, companies are failing to leverage AI because many executives have forgotten that technology only works through people, and people work best when they feel cared for. This also means that wise leaders can gain an AI advantage by leaning into connection. In particular, the following actions can help: Co-create AI strategies. In many firms, failed AI rollouts are a masterclass in unempathetic behavior. Leaders announce massive, unrealistic goals, giving employees little guidance on how to meet them and even less assurance about their own role in a post-AI landscape. This is a sure-fire way to slow adoption, encourage workslop, and raise resistance. Empathy means replacing announcements with conversations, and answers with questions. And that entails talking with employees about how AI can help them focus on the parts of their work that bring them meaning. This strategy, often called “augmentation” (as opposed to automation), is likely to provide more value over the long term. But it can also boost AI adoption now, by bringing leaders and employees into a collaborative mindset and effectively leveraging everyone’s knowledge. Co-creating AI strategies also provides leaders an opportunity to create “procedural justice” around new technologies. People are more likely to accept policies when decision-makers communicate clearly and listen to everyone involved. In other words, empathy not only allows for more and better inputs into AI rollouts but also makes it more likely that workers will come on board. Focus on the middle layer. To successfully roll out AI, senior leaders must understand the importance of empathy, but they can’t deliver it alone. In 2020, Businessolver found that only 10% of employees believed managers had the most impact on workplace empathy. By 2025, that percentage had grown to 38%. Frontline managers have become the stewards of empathic—and unempathic—cultures, and yet this “middle layer” often doesn’t receive the training in soft skills that senior executives do. This creates a gap between the values that leaders desire and the reality that employees face on the ground. Thankfully, research makes it clear that empathy can be taught, and new programs allow for that to be done at scale. Zurich Insurance, for example, recently trained thousands of claim workers in empathic communication, improving customer experience and loyalty. Similar “empathy gyms” are now being rolled out for frontline managers, allowing them to practice listening, communication, and feedback. These skills are evergreen, and especially important to navigating AI disruption. Remember that technology is human. AI rollouts require precise strategies but also reflect a deeper philosophy around what technology is for. Many leaders’ first instinct has been to replace people with machines, through workplace automation and many products that use AI to substitute for therapists, coaches, and friends. This extractive vision can dull human creativity and atrophy social skills, damaging the pieces of life that many of us value most. But there is always an alternative. Instead of replacing people, technologies can be built to connect them more deeply. I call this “AI for interdependence,” and it can be enabled, for instance, through chatbots that nudge people to reach out to one another or teach them listening skills. Workplaces can use AI for interdependence as well: for instance, by suggesting collaboration between workers with complementary skills or identifying common ground in the face of conflict. . . . These use cases already have research support and offer immense potential to workplaces. They become even more likely to help if you enable interdependence in your AI strategies. The fastest, most-effective path to using technology well involves working with the people you’re asking to use it, not around them.

Empathetic Leadership Can Make or Break AI Adoption

Summary.

Partner Center