Your Next AI Hire Should Change Minds, Not Just Use Tools

A few months ago, while working with a client in Europe, I came across a peculiar use of AI that I am sure we have all done.

A member of their staff had three bits of information they needed to send to a colleague. They put them into an AI tool, which turned them into a formal email with a greeting, three paragraphs and a polite closing. The colleague received the email, saw the length, decided they didn’t have time to read it and asked AI for a summary.

Back came the three points.

Both could reasonably say AI had saved them time. On the surface that organization could even say they are implementing AI effectively. Yet between them, they had added two steps to a workflow that weren’t even needed. The AI wasn’t even needed.

When we discussed it, we ended up talking about why staff felt they had to write formal emails to one another. Simply promoting a culture where people don’t mind receiving three bullet points would have solved the problem.

I’ve thought about that conversation a lot since. It gets at something I think organizations are missing: someone who understands AI well enough to recognize when we would be better off without it.

The Data Is Backing This Up

McKinsey’s 2026 State of AI survey found that 80% of respondents felt AI had improved their individual productivity. Just 37% attributed any positive impact on their organization’s earnings before interest and tax to AI.

Of course, benefits can take time to show up in the accounts. But still, the difference should make leaders cautious about treating individual enthusiasm as evidence that the whole organization is improving due to individual AI use.

In Microsoft’s 2026 Work Trend Index survey of 20,000 workers who used AI, 58% said they were producing work they couldn’t have produced a year earlier. Among its most advanced users, that rose to 80%. I have seen this in my company’s work and in that of my clients. But meaningful benefits have to move beyond individual gains.