Dr. Gleb Tsipursky studies the future of work. The former Ohio State University professor is the author of a book entitled “The Psychology of AI Adoption at Work: From Resistance to Results.” He’s also chief executive of a company called Disaster Avoidance Experts, which promises its customers that it can help them build their own AI solutions.
This interview is lightly edited for time and clarity.
DN: Can you give me a range of what are the different issues that managers have to think about when they think about AI in the workplace?
Gleb Tsipursky: The crucial thing that they need to realize is that the typical approach to technology adoption does not work for AI and I see managers all the time using the typical approach to technology adoption.
Here's the typical approach. The typical challenge with technology adoption, like adopting a new CRM or a sales enablement software, is that it's difficult for people to learn. It takes a lot of effort for them to learn which buttons to click and so on, how to adapt their workflows. They need to change their habits and they need very clear, top-down instructions that are called SOPs, standard operating procedures, for how to use this tool. That's what they're used to and they're trying to do the same thing with AI tools.
But that really doesn't work, because the resistance to people adopting AI tools is not coming from the same resistance that came from technology adoption in general, which is people have a lot of trouble learning new things. They find it's a lot of hassle. They have a lot of trouble changing their habits to have new patterns and workflows.
With AI tools, it's very different. These tools are very easy to learn. You talk to them, use natural language programming to just talk to them and ask them to do what you want them to do. It's very different than previous technologies, much easier to learn. But it's also very bad in terms of standard operating procedures, because they need to adapt AI to each individual person's workflows.
It's a very flexible, very customizable tool nd each individual person must be individually engaged and motivated to figure out how to use it for themselves.
So that's very different and the resistance to change, one is the nature of the tool and how each person needs to be individually motivated and engaged with it. The source of resistance is very different. The source of resistance comes because people are not excited about using AI tools.
There was a recent peer research survey showing that something like 52 % of respondents are more anxious than excited about AI tools. Only 9% of respondents are more excited than anxious. Guess which camp the executives fall into and which camp employees fall into? The executives are thinking, this is great. This will save a company a lot of money. Employees are thinking, well, I'm training a tool to replace me and so there's a lot of fear.
DN: Is that where the anxiety is all about? It's the fact that this might replace me in a couple of years?
GT: There are three sources of anxiety and one of them is definitely this.
So I talk about that in my book. I talk about how one of the three sources of emotional blockers is the fear and anxiety about job loss. That's the psychographic profile I call the AI alarms and that's based on over 600 focus group responses, many thousands of survey responses, over 100 consulting projects.
The second group that we need to talk about are going to be people who have a sense of identity threat. That's the emotional blocker.
DN: What's the identity threat about?
GT: Well, people don't feel threatened to their identity by using a CRM to look up client information instead of various spreadsheets. But they feel threatened to who they are as a professional by an AI tool that writes great sales outreach sheets, that creates great images, that creates great articles, that does really great financial analysis. They feel threatened to who they are as a professional, to their sense of professional identity and a lot of people get a sense of meaning from their work. So they feel threatened to their sense of meaning. So that identity threat is a second big block.
I call the second group, the pragmatic resistors, the first group is the AI alarmists. They don't adopt AI. They're very resistant. The third group does adopt AI, but they're reluctant to talk about it because of shame.
There's a lot of shame and social stigma applied to people who talk about using AI in their work to do sales emails, to do marketing copy, to do their image creation, article drafting, all of this stuff that people do with white collar work, financial analysis reports. I call this group the reluctant adopters. They are adopting it, but they're reluctant to talk about it.
It doesn't show up in the company productivity. Individually, people are much more productive, but they finish their tasks and then they spend their time playing around on Facebook or listening to podcasts and radio shows that they like and so this is the reluctant adopters.
As a result of all of these three blockers, what happens is what we've seen with a recent MIT study, that 95% of AI pilots fail to scale. They fail to show return on investment in terms of resources invested and so leaders are seeing these problems, but they're not getting why these are problems. They're not understanding people's psychology.
DN: I want to go back to the question of why are people reluctant to talk about it? Is there that notion that, hey, I asked a machine to do my work and therefore it looks like I'm a slacker? If I admit that a machine was doing the work that maybe two years ago, I would have been doing.
GT: You're exactly right, Doug.
Let me share a story. I was doing a presentation for a group of executives. This is a peer group that was in Pittsburgh about two weeks ago and I was talking to them about this shame on social, and one of them shared a story.
He said, oh, I know exactly what you're talking about. But recently his VP of sales was giving a presentation to all the other executives. This was a sizable company, several hundred people, so giving a presentation to all the other executives about the top client of last year, all the stories, all the patterns, what they're going to do and how they're going to do things differently because of what they learned.
It was a great presentation and he does it every year. Then after the presentation, which went great, all the other executives loved it, he approached privately the CEO who was in the group that I was presenting and told him, hey, you know that presentation, it usually takes me about three days to put together. But this year I gave it to my executive assistant and he put it together in one and a half hours. I just checked it and it was great. That was excellent. But I don't want you to tell any of the other executives because they'll think I'm lazy and they'll think I'm cheating.
That's the attitude that pervades at the top of the organization. Of course it percolates down and so that's what people feel, that cheating, that laziness. That's a very powerful social stigma. Organizations are on social reputation, social capital and if somebody thinks, oh, the guy's a cheater or the girl's lazy, well, that really ruins your career.
DN: You're making the case that AI is here to stay and that businesses really can't afford to ignore this. What are your concerns about AI? What are your reservations about AI and the way that it is developing in our society?
GT: My main reservation is concerns about young people and job loss.
There's a recent study out from Stanford showing that in the most AI-exposed industries, compared to the least exposed, in June 2025, the most AI-exposed industries young people, age 22 to 25, had 15% lower employment prospects than people in the least AI-exposed industries. They updated this data in July 2026 and they found that gap had risen to 19%. So 19% of the most AI-exposed industries compared to the least AI-exposed industries. And, of course, that gap will keep rising because AI right now is the worst it will ever be.
So my concern is young people and how do you address the talent pipeline that will be harmed if young people are not getting jobs in the industries that are most exposed to AI.
I think the crucial thing to think about is that AI will change over time and the technology will change. It will get smarter and better and better. The thing that will not change is human psychology. Human psychology and its response to AI will keep being persistently the same.
