5 AI adoption strategies that recognize one-size-fits-all won’t work

With artificial intelligence becoming more deeply integrated within both everyday life and work, employers are investing heavily in AI tools and employee training to work with the tech.

However, new research suggests that using a “one-size-fits-all” strategy can hide a fundamental truth: Employees don’t respond to AI the same way.

In fact, the research study from Shrihari Sridhar, a senior associate dean at Texas A&M University’s Mays Business School, and Huachao Gao, an associate professor of marketing at Mays, concludes that AI adoption has often been treated as if employees share a single attitude toward the emerging technology. Their findings suggest quite the opposite: People experience AI in markedly different ways depending on their individual perspectives and the context in which they’re using it—a concept they call “AI-heterogeneity.”

“What we’re seeing is that leaders are almost mandating that employees need to be AI-ready,” Sridhar says, and that’s one reason they’re encouraging employees to become familiar with AI. Another could be market pressure. A third possibility is they don’t yet know exactly how to use AI to transform their business, but they believe getting employees to use more of it will help drive that change, Sridhar says.

Employers that are mandating AI usage may be creating a “compliance rather than adoption” situation, Sridhar says, adding that this perspective is driven by the Texas A&M research that drew on a nationally representative survey of 2,144 U.S. adults.

Rather than falling neatly into pro- or anti-AI camps, he explains, many survey respondents expressed mixed feelings about the technology, seeing it as both beneficial and threatening.

“Those attitudes often shifted depending on the task at hand. Someone might enthusiastically use AI for one task while distrusting it for another, suggesting adoption isn’t a simple yes-or-no decision,” he says.

The survey also identified what the researchers call the “skepticism-usage” paradox: Some of the people most concerned about AI are among its most frequent users.

“The most anxious people are often the ones using more AI,” Sridhar said. “They’re using it, and they’re anxious about it at the same time.”

He believes this can be attributed to several factors, and he offers ways employers can resolve this AI outcome. The findings, in fact, led the researchers to reframe AI adoption as a segmentation challenge, rather than simply a technology challenge.

Sridhar suggests that rather than asking how to get more people to use AI, employers should ask how different groups experience AI and tailor their approaches accordingly in the following ways:

Audit your adoption dashboard

According to Sridhar, most organizations measure AI adoption by usage: logins, prompts, active users. Their research suggests usage alone can mislead, because the employees using AI most heavily are often the most anxious about it. “For an HR leader, the practical move is to measure confidence alongside usage,” he says. “High use with low confidence is not a success story; it is a retention and burnout risk wearing a productivity costume.

Segment your workforce the way marketers segment customers

Employees hold “a portfolio of attitudes” that shift with the task and with what the technology means for their livelihood. For example, a senior attorney whose value rests on judgment and a junior analyst whose daily tasks are exactly what AI does well are living through different events. “One all-hands email treats them as the same person, and it will miss them both,” he says.

Start with the ‘Monday morning’ problem, not the strategic vision

Sridhar says the most effective adoption message is not “AI is the future.” It is “Here is the most tedious part of your week, and here is how to hand it off.” To Sridhar, sustained behavior change comes from usefulness people can feel in their own role, not from enthusiasm for a technology in general.

Say the honest thing out loud

Employee anxiety about AI is a reasonable response to a real situation, and employees know it, Sridhar says, adding that messages that acknowledge the legitimate concern before making the ask are more credible than pure cheerleading. “HR is often the only function positioned to insist on that honesty in official communication,” he says.

Train before you measure

Sridhar says that if an employer raises the expectation of AI fluency faster than it builds the skill, employees default to shallow, anxious use that looks productive on a dashboard while confidence quietly erodes. The better path, he says, is to “close the preparation gap first” via role-specific training. The adoption numbers that follow will mean something, he says.

“Start from the workflow. Start from the job to be done,” Sridhar concludes. “Seeing how the job was done earlier, how you can do it better and where you can bring in AI is a much more productive, calming conversation.”