9 mistakes leaders make with AI strategies

At the 2025 MIT Sloan CIO Symposium, many high-tech and business leaders expressed the same complaints. This is about AI initiatives that do not provide business value, pilot projects that have not led to production, and the ongoing struggle to understand what is at stake.

So, we asked these leaders and AI experts in the room to share the lessons we learned by asking important questions: What are the common mistakes that organizations still make when shaping their AI strategies?

Their answers revealed the patterns. “The drooping fruit – it's not as low as we think,” said George Westerman, senior lecturer at MIT Sloan, warning of overestimating AI's capabilities. However, the challenges are far deeper than technical limitations.

Monica Caldas, CIO vice president and CIO at Liberty Mutual Insurance, has rethinked how the organization operates, highlighting the need for cross-working teams and cultural change management.

McKinsey's partner Hannah Mayer made observations that spoke about the speed of change. “Employees are three times more willing to use AI than leaders expect,” she pointed to disagreement among executives as a bottleneck that holds the organization down.

Watch the video and listen to advice on avoiding these important mistakes.

Setting unrealistic expectations By overestimating current AI tool capabilities.
We are missing out on opportunities for change By treating AI like another software tool.
Stuck in pilot mode And production expansion will be shorter.
I encounter the executive's hesitation I'll move forward slowly.
Forget the human factor And they're focusing too much on technology, not people.
Underestimating security risks And he didn't build resilience.

These AI strategy errors can be avoided once you know what to look for. Watch this video and learn from your peers' experiences and avoid the pitfalls that organizations are sacrificing time, money and competitive advantage.