The Hardest Part of AI Transformation May Be Redesigning What Already Works

At a CXO roundtable in San Francisco, one question sparked a particularly interesting discussion: Which departments or roles are most resistant to AI transformation?

IT teams and experienced technical professionals came up in the conversation.

It made me think about a deeper organizational paradox: the people most successful in the existing system may have the strongest incentives to preserve it.

Not necessarily because they oppose AI, but because their expertise, credibility, and career success have been built around workflows that AI is beginning to redefine.

This leads to another question we discussed: How should performance management evolve in an AI-native organization?

For decades, companies have rewarded individual output, functional expertise, and efficiency within established processes.

But what happens when AI can generate more code, complete more tasks, and produce more deliverables—without necessarily creating more enterprise value?

Individual efficiency is not enterprise productivity.

The real challenge is to rethink what good performance means.

• Are we rewarding people for completing tasks faster, or for redesigning workflows that create greater value?
• Are we evaluating individual contributions, or the organizational capabilities they help build?
• Are our performance systems encouraging transformation—or quietly reinforcing the status quo?

Great exchanging perspectives with David Murray, Kim Homenock (Newton), Niyati Goel, Angela Rui, and fellow leaders at the roundtable.

Perhaps the biggest barrier to AI transformation isn’t resistance to new technology. It’s continuing to reward success by yesterday’s rules.