Hi Reader,

Last week I was presenting at WebExpo in Prague — talking about the good and bad of human–AI collaboration.

After my talk, one of the attendees came up to me and asked: "How do I know if my team is using AI in the right way?"

It's an interesting question because, as I've written before, there is a right and a wrong way of using AI.

​Research suggests that when AI comes first — when people receive a summary, a draft, or an insight before doing their own thinking — their brains disengage more quickly. They remember less and start losing ownership of the work.

But when people think first and use AI second, their brains stay much more engaged.

There's another thing I've been writing about as well.

When teams use AI regularly, they start adopting AI language, AI ways of framing things, and AI terminology without even realizing it.

And that can create artificial certainty.

AI can make us feel certain about something that may not actually be true. It can create a false sense of confidence.

A new paper published in Organization Science studied two urban planning teams using the same AI simulation tool. Both teams were experienced. The difference was in how they worked with AI and then presented the AI's outputs to stakeholders.

One team increasingly presented the outputs as if they were predictions of the future rather than scenarios to explore. The other framed them as exploratory — "this suggests a possibility."

In the first team, the model became so authoritative that stakeholders increasingly relied on it and sometimes bypassed the planners' interpretation. The AI output looked so complete, so structured, and so certain that there seemed to be little left to evaluate. There's an irony in that: the experts who leaned hardest into the AI's authority ended up undermining their own.

The second team preserved their authority. Their stakeholders stayed engaged, stayed critical, and stayed curious.

The researchers called the first dynamic "artificial certainty" — the illusion that complex, inherently unpredictable futures are definitively knowable. When information is presented in a confident, structured tone, people are more likely to accept it as truth. When the same information is framed as a probability or a possibility, they evaluate it more critically.

The danger isn't that AI is wrong. It's that it looks so right that nobody questions it anymore.

So what does that mean for you and your team? It means team members using AI need to build the discipline of questioning AI outputs, no matter how certain those outputs sound. Critical thinking becomes more important, not less.

So what does that mean for you and your team? It means team members using AI need to build the discipline of questioning AI outputs, no matter how certain those outputs sound. Critical thinking becomes more important, not less. If you'd like to watch my talk from Prague, here's the link. Partnering with AI: Building future-ready teams

See you next Thursday.

Daria

P.S. If you'd like to discuss how AI is influencing your team — the risks, the opportunities, and where it might be taking over your team's thinking — book a call with me.

Check out more of our work at...

​ Linkedin ​Connect ​ →​

​ Youtube ​Subscribe ​→​

​ My book ​Read ​→​