New research is highlighting that American employees are using AI tools and reporting they don't like them. That's not a contradiction - it's a measurement problem, and one that can no doubt be extrapolated into all geographies.

Usage and satisfaction tell you different things. People are using AI tools because their employers have deployed them (in some cases, mandating their use), because the efficiency gains are real in specific tasks, or because professional norms are shifting. Employees are reporting dissatisfaction because the experience is often inconsistent, the outputs require significant rework, and the tools are being applied to problems they're not suited to.

The response from vendors has mostly been to improve the tools. That's the wrong first move. The more immediate problem is that organisations are deploying capability ahead of capability-building. The technology is in the hands of the workforce before the workforce knows how to use it well.

Singapore has recognised this clearly enough to treat AI workforce literacy as a national policy priority. The EU has mandated it through Article 4. Most countries, however, are leaving it largely to the market - which means leaving it to individual organisations, many of which have no clear plan.

The satisfaction data isn't a signal that AI doesn't work. It's a signal that adoption without enablement produces exactly the outcomes you'd expect.

#AI #WorkforceTransformation #FutureOfWork