Employees do not encounter AI in isolation. Their response to the technology is shaped by the social environment surrounding them, including trust, perceived support, psychological safety and expectations about what AI means for their future. New evidence on workplace empathy illustrates why AI adoption should be understood as a human performance phenomenon as well as a technology outcome.
The Technology Is Only Part of the Environment
Two employees can receive access to the same AI system, attend the same training and perform broadly similar roles, yet respond to the technology very differently. One experiments, asks questions and gradually incorporates AI into increasingly sophisticated work. Another uses it cautiously, avoids situations where mistakes might become visible and remains uncertain about whether the technology represents an opportunity or a threat.
Differences of this kind are often explained through individual attitudes towards technology. Some people are assumed to be naturally curious, others resistant to change. Human Performance Intelligence™ suggests a broader interpretation. People encounter technology within an existing performance environment, and that environment influences the cognitive, emotional and behavioural resources they bring to the interaction.
New findings from Businessolver’s 2026 State of Workplace Empathy AI Special Report provide an interesting illustration. Among employees who described their workplace as empathetic, 87% reported receiving adequate AI training, compared with 33% among those describing their culture as toxic. Employees in empathetic environments were also substantially more likely to report optimism about their future with AI and a greater sense of agency over their work. The research demonstrates association rather than causation, but the consistency of the differences across AI-related measures makes the relationship worth examining.
People Interpret Technology Through Their Environment
The significance of these findings extends beyond empathy itself. Human beings do not respond to organizational change solely according to the objective characteristics of the change. They interpret what it means for them. AI can represent an opportunity to increase capability, reduce tedious work and expand professional possibilities, while simultaneously creating concerns about competence, status, job security and future relevance.
The social environment provides much of the information through which those meanings are constructed. Employees observe what leaders say about AI, what managers reward, how mistakes are treated and whether questions can be asked without reputational cost. They notice whether the organization invests in helping them develop or simply expects them to adapt. Long before an employee decides how extensively to experiment with a new AI system, these signals have begun shaping whether engagement with the technology feels safe, worthwhile and personally controllable.
This helps explain why formal access to technology can produce such different behavioural outcomes. The tool may be identical, but the human performance conditions surrounding its use are not.
Social Conditions Can Influence Cognitive Performance
This relationship is particularly important within the HPI architecture because Social and Interpersonal Conditions sit upstream of several other dimensions of human performance. The quality of relationships, leadership and psychological safety can influence stress responses, motivation, cognitive availability and ultimately the capacity to adapt.
An employee who believes that AI experimentation is supported can devote attention to learning how the technology works, evaluating its limitations and developing more effective ways of using it. An employee who is simultaneously concerned about being replaced, appearing incompetent or making a visible mistake is operating under a different cognitive and emotional load. Part of their available capacity is directed towards managing uncertainty and threat rather than towards exploration and learning.
Businessolver’s findings are consistent with this interpretation. Thirty-nine percent of employees reported concern about what AI meant for their future at the organization, while 31% worried about falling behind in their ability to use AI effectively. Among employees describing toxic cultures, those concerns were substantially higher.
The implication is important: what appears to be resistance to AI may sometimes be an observable outcome of conditions elsewhere in the human performance system.
Adoption Metrics Can Hide the Mechanism
Organizations understandably measure AI adoption through indicators such as logins, frequency of use, training completion or the number of active users. These measures describe behaviour, but they provide limited information about what is producing it.
Two teams with identical adoption rates may be experiencing very different realities. One may be experimenting confidently and discovering increasingly valuable applications. Another may be using AI because usage is expected while remaining reluctant to rely on it for consequential work. Conversely, low usage may reflect anxiety or insufficient support, but it may also represent informed judgement that the technology performs poorly for particular tasks.
Understanding human performance therefore requires moving beneath the behavioural indicator to examine the conditions generating it. Perceived agency, psychological safety, workload, confidence, managerial support and the meaning employees attach to AI can all influence whether technological capability becomes actual human behaviour.
This is one reason HPI treats the dimensions of human performance as an interconnected system rather than independent variables. Social conditions can alter stress and motivation; those changes affect available cognitive capacity; and together they influence whether people have the resources and willingness required to adapt.
AI Adoption Is a Human Performance Outcome
The Businessolver research should not be interpreted as evidence that empathy alone causes AI adoption. The relationship is likely to involve multiple mechanisms, including differences in training, management quality, communication and broader organizational culture. That complexity is precisely what makes the findings interesting.
AI adoption emerges from an interaction between technological capability and human conditions. Organizations can provide access to sophisticated tools, but the value ultimately created depends on whether people have the cognitive capacity, motivation, confidence and social environment required to learn how to use them effectively.
This leads to a broader proposition for Human Performance Intelligence™. AI does not enter a neutral human system. It enters organizations in which people already experience particular levels of trust, stress, autonomy, motivation and psychological safety, and those conditions influence how the technology is interpreted and used.
For organizations seeking to understand why AI produces very different outcomes across teams, functions or locations, the explanation may therefore begin well before anyone opens the tool. The quality of the surrounding human performance environment can determine whether technological potential becomes experimentation, learning and adaptation, or remains capability that the organization possesses but never fully converts into performance.