62% of workers surveyed in 32 countries say AI saved them time last year. Yet only 40% of Americans think its benefits outweigh the risk of job losses. Use is growing faster than trust, and that gap is what weighs most on adoption in companies.
The facts
- Only 11% of corporate communicators consider their ethics policies adequate for AI.
- 44% of Americans oppose the construction of new data centres, according to an Ipsos survey from June 2026.
- Media coverage judged favourable to AI fell to 38% in June 2026, from 44% a year earlier.
Why does the gap between use and trust slow adoption?
The paradox measured by Ipsos is the one found in most organisations: teams use AI, often more than they admit, while doubting its effects on their jobs and their company's ability to use it ethically. When only 11% of communicators consider their internal policies up to the task, the message sent to teams is clear: the tool is moving faster than the framework.
This gap produces very concrete effects: token adoption in some departments, hidden use in others, and open resistance as soon as an official project is announced. An AI rollout that ignores the question of trust sooner or later runs into this internal resistance, and the climate of opinion, measurable in the growing distrust of data centres or the reversal of media coverage, will only amplify it.
AIxH's view
That is why our AI training courses address trust as much as hands-on use of the tools: explaining what AI does and does not do, setting a clear framework for use, and giving teams the means to judge for themselves. A team that understands the tool no longer needs to fear it, or to use it in secret.