Is Your Organization Messing Up AI Adoption?
smashingmagazine.com ● Covered by 8 sources
Most companies are rolling out AI like a software license, not a change process. That’s why staff get nervous, use shadow AI, and the whole thing slips out of control.
Based on reporting by smashingmagazine.com — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Paul Boag says the way many organizations are adopting AI looks less like a plan than a mess. His case is simple: this feels a lot like the digital transformation failures he wrote about in 2014, and the same mistakes are back in new clothes.
The core error is treating AI as a technology project instead of a people problem. In the digital transformation years, companies pushed out CMSs, CRMs, intranets, websites, and apps while paying too little attention to how people would actually use them. Now management is buying into the hype, IT teams are rolling out Copilot Studio licenses across the company, and that is often being treated as job done.
Boag argues that staff have little reason to be enthusiastic. AI is mainly sold internally as a productivity boost for the company, not as a benefit for the people using it. They are still expected to work the same hours, do more, and somehow feel grateful about it. On top of that, many are worried about their jobs, which makes cheerful rollout emails from management sound a bit thin.
That means adoption needs incentives, time, and support. Boag says organizations should help people find the parts of their work they dislike and show how AI can make those tasks easier, but also think about career progression, compensation, or even shorter working hours. Training has to be role-specific, not just a generic session, and the best internal support often comes from the people already experimenting with AI rather than the ones digging in their heels.
He also warns that people will not experiment if every mistake feels dangerous. Some organizations still demand the same KPIs, then complain when staff use AI in clumsy ways, or even police token use just weeks after pushing adoption. Governance matters, but not the kind that comes in pages of rules. And when IT locks everything down, the result can be shadow AI: staff opening personal ChatGPT or Claude accounts on phones or home laptops because the official tools cannot do the job. Boag’s point is blunt. If organizations keep focusing on control first, they will miss the real challenge entirely.
My take — AI-written commentary, not fact-checked reporting
This is the usual corporate trick: buy the tool, skip the change, then act surprised when nobody uses it properly. Open vs closed models is almost beside the point if the policy is fear, bureaucracy, and a training deck nobody remembers by lunch. The real AI strategy is still the boring one — trust people enough to let them work, then fix the mess they uncover.
Read more about this at: smashingmagazine.com