The Psychological Costs of Adopting AI – Executive Summary
HBR, by Guy Champniss, published May 2026
1. Summary
"The Psychological Costs of Adopting AI" (HBR, 2026) argues that AI adoption stalls because working with AI tools puts psychological strain on the people who are supposed to use them. Six hidden costs that erode motivation are identified: skill loss, a diminished sense of autonomy, growing doubt about one's own competence, reduced social contact with colleagues, a loss of credibility in front of peers, and a threat to professional identity. These costs rarely show up in usage dashboards, which is why organizations keep reporting healthy adoption numbers while the underlying experience of work gets worse. Early-career employees are the group that gets hit hardest, since they are handing over the tasks that would otherwise build their judgment. The risk described is a generation that never develops core skills because AI did the work first.
2. Relevance for CIOs
For CIOs, the practical consequence is that adoption metrics alone give a misleading picture. Tracking license activity and prompt volume says nothing about whether people feel confident and capable next to the tools, so confidence and perceived competence belong in the measurement set alongside usage. The six cost categories also give a design brief: build in deliberate "think first" moments before employees reach for AI on complex or strategic work, so that judgment still gets exercised where it matters. Because credibility and identity are part of the picture, AI use has to be visible and normal across the organization; where people hide it or feel sheepish about it, the culture has not caught up with the technology. Since early-career staff carry the highest exposure, development paths need rethinking so that juniors still earn the reps that seniority is built on.
3. Thoughts from Sonny Claussen (CEO Bee360) about this Article
Sonny sees the pattern in his own team: some people are energized, some are anxious, and most are both at once. He points to Bee360's own shifts as the setting for that tension, from rebuilding the website out of a static setup into something that adapts to every market the company operates in, to Professional Services colleagues now working with product knowledge that reaches them in the right context, the right language, and the right moment. For him this was the most uncomfortable read of the month, and a useful counterweight to every AI story that only counts what went right. His conclusion is that the gap between AI opportunity and AI reality is a human and organizational problem, and that it cannot be closed by moving faster.