If you ask board members or a CEO how they feel about AI, the answer is likely to be peppered with words like efficiency, cost savings and innovation. The enthusiasm can be heard in press releases, investor calls and company town halls.
Employees are often much more wary of the most human-adjacent technology of their lifetimes.
A recent survey of 2,400 Gen Z Workers found that 48% think the risks of AI in the workforce outweigh its benefits. The survey was conducted by Gallup in partnership with Making Caring Common, a program of the Harvard Graduate School of Education, and the Walton Family Foundation.
Meanwhile, a 2025 Pew Research report found a sharp divide between AI experts and the public over AI’s impact on work. While 73% of 1,013 AI experts surveyed said AI would have a positive or somewhat positive effect on how people would do their jobs during the next 20 years, just 23% of 5,410 U.S. adults surveyed agreed.
CIOs ignore that disconnect at their own risk. They are expected to lead the charge on enterprise AI adoption and deliver business value. Success will depend in part on bridging the gap, which will necessitate closing the disconnect between leadership’s full-throated enthusiasm and employee skepticism.
Understanding the negative sentiment
Employee concerns regarding AI abound. They are worried about how it is changing the way they work, the safety of their jobs and its environmental impact.
“You can’t put your head in the sand and pretend these [concerns] don’t exist, because they do. They’re in the headlines every day. It’s on TV every day, what you watch, what you hear, what you listen to,” said Thomas Phelps, CIO and senior vice president of corporate strategy at Laserfiche.
Job anxiety is one of employees’ biggest concerns. Efficiency gains are AI’s biggest selling point for executives and boards. The “do more with less” ethos is an exciting prospect for companies.
At the employee level, however, this can mean the expectation to do more work. They are the ones doing more with less.
We even have a name for the mental fatigue people are feeling as they use and oversee more and more AI tools at work: “AI brain fry.”
Workplace surveillance is adding to those concerns. At some companies, employees are subject to not only their bosses’ oversight but also AI surveillance tools that track how they work. Meta is using the data it captures with these tools to train its AI models, Reuters reported in April. Working in that kind of environment can easily feel as if you are training your eventual AI replacement.
Not every company excited about AI is using it for workplace surveillance and cutting head count. There are even arguments that AI ultimately create more jobs than it replaces. But that doesn’t change the reality employees see today: mass layoffs and the “low-fire, low-hire” job market.
Whether or not AI is directly responsible for layoffs and slowed junior hiring, job anxiety is high. Just 22% of workers around the world feel that their jobs are safe from elimination, according to a survey from ADP Research of more than 39,000 working adults.
Environmental concerns are growing. In addition to job anxiety, people are increasingly worried about the environmental impact of AI. A Gallup poll of 1,000 adults found that 70% of Americans are against AI data center construction in their local communities.
Diagnosing and addressing AI pushback
Given the high levels of job anxiety, understanding true AI sentiment within an enterprise is not an easy task for CIOs.
“My job as CIO and having the responsibility for AI means that most people are probably not going to come to me and say, ‘Darren, I don’t like this,’ ” said Darren Cassidy, CIO at Sitecore.
Usage data provides an early signal of employee sentiment. Usage is one of the baseline metrics for taking the temperature of AI enthusiasm — or lack thereof — among employees. Are employees leveraging the tools available to them?
“If I see a drop in that engagement … could it be sentiment? Could it be they don’t feel there’s value in it?” asked Galen Counselman, senior vice president and CIO at PenAir Credit Union. “That’s at least an indicator for me to do some digging into it.”
At Laserfiche, Phelps said he has seen varying levels of usage across tools, which can shape decisions about which tools are making a positive impact.
“Copilot is a tool that we’ve licensed in a limited capacity. We’re not finding a lot of usage,” he said. “But OpenAI and Anthropic tools, we’re finding a lot more usage. Gemini, we’re finding more adoption. From an IT perspective, we want our people to know that we’re listening to them. We’re not going to license tools for everyone. We want to do it in the right way and pick the tools that best serve them.”
Usage metrics are helpful, but they don’t tell the whole story. CIOs can get more insight into how workers feel about AI through conversations with their own teams, leaders of other business units and employees across the company.
“We would have meetings and whatnot, and I would just ask some of our employees: ‘Hey, what’s your thoughts on AI?’ And frequently I would get, ‘Scary,'” Counselman said.
Last year, PenAir began holding meetings every other week that anyone at the credit union can join with the intent of creating a platform for Copilot users to have open discussions. On any given week, roughly 30 to 40 people join for wide-ranging conversations that focus on prompts, how people are using the tools, and issues like AI water usage.
“I found that’s been an effective way to take away some of the fear of it and help make people more comfortable around the tools,” Counselman said.
Sitecore runs quarterly surveys to get a sense of how employees feel about the technology they are using. Do they feel they have the right tools, training and support?
Cassidy noted that as the company works through AI adoption, it has made changes to its AI metrics.
“We gave everybody a global KPI about using AI. That was, I wouldn’t say it was not popular, but it was not met with hallelujahs all around,” he said.
Instead, Sitecore has moved to tracking a series of different metrics.
“It’s been a learning curve, trying to get away from that financial success metric and move to ‘What does a productivity metric look like? What does a sentiment metric look like?'” Cassidy explained.
Employee advocates can help build support for AI initiatives. If CIOs have a clear understanding of where employees are unhappy, they can’t fix the issue or series of issues themselves. They may be able to make significant inroads with employees by identifying and empowering people across teams who are interested in using AI. Give them tools. Elicit their feedback. Listen to what they say about the good and the bad. Their voices can influence and support an enterprise’s AI strategy.
“It’s not just me saying, ‘Hey, you ought to use this stuff,'” Counselman said. “We have 40, 50, 60 people that are also saying, ‘Hey, yeah, this is great. It’s helping me out.'”
But finding those people and getting their honest thoughts requires the right kind of company culture, one not shot through with fear of job loss.
“You want to create a culture where employees feel comfortable raising their hand, sharing their concerns and knowing that feedback is going to be taken in terms of actions or initiatives that can help them versus negatively impacting them,” Phelps said.
The risks of ignoring employee concerns
“Get onboard or get left behind” is a well-worn mantra in the AI space. Workers may very well get left behind, but if CIOs and enterprises opt to ignore their employees’ concerns, they risk being the ones left behind.
“If CIOs aren’t paying attention to this and they’re just railroading this through and just forcing the issue, I think they have a high chance of having some big culture and morale issues on their hands, which could turn into losing a lot of good talent,” Counselman said.
Even if enterprises don’t lose talent, CIOs could find themselves in an unenviable position of championing a technology people aren’t actually using.
“Without engaging employees, without recognizing that change is always challenging in many, many organizations, you’re going to end up in a situation where you’re overspending on AI but find employees are not using it,” Phelps warned.
With increased pressure to move from pilot projects to actual ROI, CIOs will face tougher questions from their C-suite peers and boards. CIOs need to justify AI spending with tangible results while ensuring strong governance.
“If a CIO can’t do that, the CEO is just going to replace them,” Cassidy said.
Communicating with leadership and employees
If CIOs want to be successful, they have to figure out how to deliver on board expectations without burying employees’ reluctance and fears.
“As an IT leader, you’ve got to listen to people and understand what their concerns are and figure out a way to bridge that gap, and hopefully understand what motivates individuals, what motivates teams and departments,” Phelps said.
Transparency is a valuable tool for assuring both boards and employees. To boards, CIOs can be frank about why employees are concerned and how addressing those concerns will serve the business. To employees, CIOs can talk about leadership’s overall AI goals and strategy. Is the goal to reduce head count? Is the goal to slow junior hiring?
CIOs cannot promise employees there will never be job loss, but they can be honest about the board’s intent and what that means for employees.
As Sitecore’s leaders worked to establish its AI policies, they also spent time on employee messaging. Cassidy said they talked with employees about how much the company was investing in internal AI tools and the anticipated benefits.
“We had three to four months of foundational work that we had to do, along with the PR aspect of selling this to our employees,” he added.
Navigating the communication around AI boils down to trust, according to Counselman. “Talk about the issues,” he said. “Openly talk with the board about them and then also openly talk with your team about what you’re doing and why, and work on building trust.”
Open lines of communication that run both up and down within an organization can help CIOs build a cohesive strategy for AI implementation that supports enterprise expectations without disregarding the reality for employees using the technology. Deploying AI tools without a deliberate strategy is a mistake, according to Cassidy. “If you deploy all the tools to everybody, but you have no forward thinking or planning, all you’re going to do is scare, frustrate and disempower the employees that you’re trying to put the AI in the hands of.”
Enjoyed this article? Sign up for our newsletter to receive regular insights and stay connected.