From visibility to responsibility
A few newsletters ago, we wrote about Shadow AI — tools adopted quietly because the pressure to deliver outpaced formal guidance. The central issue was visibility: what leaders couldn’t see, they couldn’t trust.

Then we talked about productivity and this feeling of hesitation that sometimes plagues larger organizations. The issue there wasn’t capability, but accountability.

Now the conversation is shifting again. AI is no longer informal, optional, or hidden. In fact, in some organizations, it’s becoming required. As AI has evolved there’s been a constant burning question about what its impact is going to be. Is it taking all the jobs? Some of the jobs? Is it overblown? Or is it underestimated? But maybe these aren’t the right questions at all. Maybe the real risk isn’t AI itself, but rather it’s failing to define what uniquely human roles look like going forward. 

Adoption without definition creates anxiety
One of our SNP coaches recently brought this to a round table discussion. Speaking about one of their coachees, they shared: 

Product managers are now required to use — and document their use of — AI in their work. This has created anxiety among some who worry their companies are getting AI to learn from them so their jobs can eventually be eliminated. So I’m working with them to identify the parts of their jobs that AI cannot do. We’re going to lean into making them masters of those skills. 

It makes sense that we’ve largely failed to consider the human role. The technology moves faster than our roles can be redefined and so the anxiety has centered around the speed at which AI is moving. Because AI’s capabilities are improving more rapidly than we can even test them, it’s difficult to foresee where exactly it will all end up. And when we don’t know what we don’t know, people construct their own narratives about replacement, surveillance, and diminishing value. But if we shift our focus off of what AI can do and onto what we can do, we regain some of our agency. (A great example of this is The Wall Street Journal recently talked about an uptick in companies hiring “storytellers”—because storytelling is a uniquely human skill.)

Fear is rarely about the tool itself. It is about unclear expectations.
Most AI adoption initiatives are framed in terms of productivity: faster outputs, lower effort, increased scale. These are real gains. But they also shift the center of gravity of work. When leaders introduce powerful tools without redefining human contribution, employees are left to infer their future value from efficiency metrics alone. That framing narrows work to pure “output” and obscures the relational and interpretive dimensions that organizations actually depend on. 

Across the past year, a story arc has emerged:

• People using AI in the shadows established hesitancy, fear, anxiety, distrust
• A focus on productivity for productivity’s sake (without emphasizing judgment or strategy) called into question purpose
• And all the while a renewed return-to-office debate and shifting policies put the spotlight on real connection and relationships, all bringing to a head the question of where do people even fit into this equation anymore? 

And the short answer is: organizations have not fully articulated what human work is for in the age of AI.

The Human Side
So as generative tools become standard, what can people do that still gives them purpose?

• Frame problems worth solving
• Exercise judgment under uncertainty
• Maintain accountability across stakeholders
• Conduct difficult conversations
• Build trust where outcomes are not guaranteed

AI can’t do these things. Only people can. 

The leadership task ahead
AI will continue to evolve. Your job isn’t to outrun it. It’s to clarify where your judgment, ownership, and trust make the difference.

Whether you lead a team or contribute as an individual, the work that sets you apart won’t be faster output. It will be clearer thinking. That responsibility doesn’t belong to leadership alone.

It belongs to everyone.