How do you Keep a Focus on People when AI Is Driving the Agenda?

Your company wants to implement new AI solutions.

Leadership teams are facing pressure from competitors. Technology teams are excited about the new capabilities AI brings. Operations leaders see greater opportunity for efficiency. And someone on the leadership team already has a platform in mind to implement AI.

Meanwhile, you may be wondering whether anyone has spoken with the customers or employees who will experience the solution.

If you are a product manager, researcher, designer or customer-experience professional, raising that concern can feel like resisting progress. But it isn't.

You can help the company invest in the right AI, in the right places, for the right reasons.

The rush toward new AI solutions is getting ahead of a confident understanding of what customers want

Gartner surveyed 1,303 leaders at organizations with more than $50 million in revenue in early 2026.

AI investment is growing much faster than the ability to show what it returns.

The people asking difficult questions may not be slowing AI progress. They may be protecting the organization from scaling the wrong solution.

Move the conversation from capability to value

AI discussions often begin with what the technology can do:

Try bringing the conversation back to the intended outcome:

1. 1What customer or employee problem are we solving?
2. 2What evidence tells us it is important?
3. 3How are people solving it today?
4. 4What becomes better for them if we add AI?
5. 5How will we know the improvement is real?

These questions do are not opposing implementation, just giving the investment a purpose.

Turn confident statements into testable assumptions

In important conversations you may hear:

These may be reasonable hypotheses but they need evidence from customers.

A constructive response is:

"That may be true. What would we need to learn from customers before we scale it?"

Qualtrics studied 7,001 consumers and found that AI customer-service agents performed relatively well on friendliness but struggled most with understanding. Understanding was also the quality most closely connected to resolving the customer's problem.

An AI experience can look polished in a demonstration and still fail in the messy reality of customers' lives.

Ask what happens when the AI is wrong

Ideal demonstrations tell us very little about how a service performs under pressure. Before approving a customer-facing AI experience, ask:

1. 1How will customers recognize that the AI may be wrong?
2. 2Can they correct or reverse its action?
3. 3Can they reach a person without beginning again?
4. 4Will their context follow them through the handoff?
5. 5Who is accountable for the outcome?
6. 6Which decisions should always require human judgment?

Failure recovery is part of the customer experience.

Ask what the dashboard might be missing

You do not have to argue against AI

You can advocate for a focused research and design phase before the organization makes a larger commitment. The most important question in the meeting may be:

THE QUESTION TO ASK

What are we assuming about customers and employees that we should validate before we build this?

Answering that question properly requires more than a quick survey or internal workshop. It may require customer interviews, employee research, journey mapping, service blueprints, prototype testing and an examination of how people, AI and backstage systems will work together.

That is where service design becomes essential.

Is your organization moving from assumption to implementation too quickly?

Stratos helps organizations identify where AI belongs in the customer experience, what must remain human and what needs to be researched before investing further.

Schedule a complimentary 15-minute AI Clarity Call. Bring one AI initiative your organization is considering, and we'll help you identify the critical human questions and assumptions surrounding it.