When the AI is the Executor
I closed my last post with a question: when the AI is the executor, what does accountable governance actually look like?
I spent the last couple of days at the YPO Corporate Governance Symposium in New York. Serious board members. Senior governance practitioners. Some of the most credentialed corporate governance faculty in the country.
Here's what I found, and what it clarified...
I closed my last post with a question: when the AI is the executor, what does accountable governance actually look like?
I spent the last couple of days at the YPO Corporate Governance Symposium in New York. Serious board members. Senior governance practitioners. Some of the most credentialed corporate governance faculty in the country.
Here's what I found, and what it clarified.
The room was still asking the risk management question.
At the highest levels of enterprise AI advisory, the dominant framing is still risk oversight. How do we monitor AI? How do we audit outputs? How do we manage the liability exposure?
That is the right question for a system that recommends. It is the wrong question for a system that executes.
The structural design question -- how do you govern a layer where AI is making decisions, not just surfacing them -- was not being asked. Which tells me the answer is not obvious yet. And that the window to develop a real one is open.
The clearest framing came from an unexpected direction.
Prof. Clifford Schorer, Co-Director of Innovation and Entrepreneurship at Columbia Business School (my alma mater), introduced a distinction I have not stopped thinking about:
CI versus AI. Creative Intelligence versus Artificial Intelligence.
His point was about education. The governance implication is sharper.
When AI is the executor, the accountability gap does not close with better monitoring frameworks. It closes with human judgment: the capacity for synthesis, for novel problem framing, for saying "the system is wrong and here is why." That is what fills the space between what an autonomous system can do and what a governing body is actually responsible for.
That is Creative Intelligence. And it is exactly what most governance frameworks are not designed to cultivate or protect.
Sonnenfeld's insight lands differently in this context.
Jeffrey Sonnenfeld's great observation: the governance failures at Enron, WorldCom, and Tyco were not structural. The boards had the right committees, the right independence ratios, the right compliance frameworks. What they lacked was a culture of honest challenge. Directors who felt safe pushing back on what the system was telling them.
Apply that to AI governance.
The board that can challenge an AI system's outputs - one that treats dissent from the algorithm as an obligation rather than a disruption - will govern autonomous AI better than any board that defers to it.
The structure matters less than the culture. It always has. But when the system being governed is making decisions at machine speed, the cost of a deferential board culture compounds dramatically.
So what does accountable AI governance actually look like?
Not a new compliance checklist. Not a better dashboard.
It looks like a Sonnenfeld board, applied to a new kind of system. Trust, candor, willingness to challenge -- directed at an AI executor rather than a CEO.
And it requires boards to actively protect the CI that makes that challenge possible. Boards that over-specify process, that reduce every decision to a framework, that optimize for compliance over synthesis, are inadvertently eroding the one thing they will need most when the AI gets it wrong.
The governance question is not structural. It never was.
The question I am now sitting with:
As AI systems move from advisor to executor, is your board developing the social capacity to challenge them, or just the frameworks to monitor them?
Reach me at arif@faris-capital.com
This thesis was originally published on LinkedIn. Join the discussion, add your thoughts, and follow for regular updates.
View on LinkedIn