AI and the human side of execution
Every company now has an AI strategy. Very few can say, a year in, what it changed. The reason is rarely the technology, and it's rarely the ambition. It's that AI plans fail the same two ways every strategy fails, built badly or never truly run, and AI adds a force most leaders have never had to manage: the technology changes what work means to the people doing it.
Built backward
Companies are building their AI plans backward: they start from the technology and go looking for uses. The result is a portfolio of pilots, each individually defensible, collectively adding up to nothing a board could call a strategy. A real AI strategy starts where every strategy starts, with a business problem worth solving, and treats AI as one candidate answer among several. When the problem leads, the pilots have somewhere to go. When the technology leads, they have somewhere to hide.
Stalled in the running
Even well-built AI plans stall in execution, for a reason few leaders see coming. When software can draft the report, run the analysis, or answer the customer, the people doing that work ask a quiet question: what is my judgment worth now? Left unanswered, that question becomes resistance no dashboard can explain, adoption that never quite arrives, and workarounds nobody admits to. The barrier isn't the tool. It's what the tool means to the person holding it.
The first failure is a strategy problem. The second is a human one, and it's the one I studied.
What the research found, and where it's going
In my doctoral research, I surveyed 505 working adults across the US to test a widely assumed story: that intelligent automation erodes people's sense of meaning at work. The data didn't support it, in either direction. Employees did not experience automation as a threat to the meaningfulness of their jobs. What the data did show is where meaning actually comes from: using a range of one's skills, owning an identifiable piece of work, doing something that visibly matters to others, all amplified by the overall climate of the workplace.
That study predates generative AI, and research since has been filling in the picture. A three-wave study of 360 employees found that AI use positively predicts wellbeing at work, running through two channels, work meaningfulness and engagement, but only when the quality of human-AI collaboration is high (Zhang, Chen, Zhang & Zhang, 2026). Work on AI implementation reaches the same shape of conclusion: HR practices and rollout choices determine whether AI fosters meaningful work (Gemmano et al., 2026), employees find work more meaningful when they interact with AI as accountable decision-makers rather than supervised operators (Hosseini, Shen & Müller-Birn, 2024), and when AI substitutes for the core of a job rather than augmenting it, engagement falls (Xu et al., 2025). Read together, the pattern is consistent: the effect of AI on people is not inherent to the technology. It is contingent on how the deployment is designed and run, which makes AI's human impact an execution question.
That's also my working hypothesis for what comes next: that AI, deployed deliberately, improves the very things that make work meaningful, more skill in use, more ownership, more visible significance, and I'm designing research to test exactly that in the generative era. The practical version of the question, the one I put to leadership teams, is this: when you decide which tasks the machine absorbs, you are deciding whether the work that remains has more meaning or less. Most companies are making that decision by accident.
Saad, S. (2021). The Influence of Intelligent Automation on Employees' Sense of Meaningfulness at Work. Doctoral dissertation, University of Dallas. · Zhang, C., Chen, S., Zhang, D., & Zhang, Q. (2026). The impact of artificial intelligence usage on affective work well-being. Behavioral Sciences, 16(5). · Gemmano, C. G., et al. (2026). Making artificial intelligence work at work. Behavioral Sciences, 16(2). · Hosseini, Z., Shen, X., & Müller-Birn, C. (2024). The soul of work: Evaluation of job meaningfulness and accountability in human-AI collaboration. Proceedings of the ACM on Human-Computer Interaction, 8. · Xu, M., et al. (2025). Examining the double-edged sword effect of AI usage on work engagement. Behavioral Sciences, 15(2).
What I do about it
I work with leadership teams on both failures: shaping AI strategies that start from the business rather than the technology, and building the adoption path that takes people's experience of the change seriously, because that's where the return actually lives. And since advice that stops at the deck is half the job, the work extends into implementation itself. This runs through my consulting practice; the way I work is here.
If your AI strategy is a year old and the business runs the way it did before, that's not a technology problem, and it's worth a conversation. I also speak on this: what AI does to the meaning of work, and why that's where AI strategies live or die. Get in touch either way.