The AI Advantage Will Be Won or Lost in the Boardroom
Your CEO cannot capture the value of agentic AI until the board has done its governance work.
McKinsey's June 2026 report, The Symbiotic Enterprise, makes the gap clear. More than 80 percent of companies use AI in at least one function, yet few report meaningful P&L impact. Fewer than 10 percent are scaling AI agents in any single function. Only 30 percent of CEOs actively oversee their organization's AI agenda.
The constraint is not the technology. It is governance.
Generative AI largely introduced organizations to machines that could answer questions, analyze information, create content, and assist people with their work. Agentic AI changes the conversation, changes the game. An AI agent can potentially access systems, draw upon proprietary information, execute workflows, interact with customers and vendors, make decisions within defined limits, and take action with decreasing levels of human intervention.
The question for a board is therefore no longer simply:
What can AI do for us?
It is also:
What are we willing to allow AI to know and do on our behalf?
That is a governance question. Robert F. Smith, Vista Equity Partners, articulated this important principle: bring AI to the enterprise rather than bringing the enterprise to AI. (see the full interview here)
For entities below the enterprise level, businesses or organization or nonprofits of any size, I believe the principle is the same:
Bring AI to the business. Don't bring the business to AI.
Before an organization begins turning over information, processes, decisions, and eventually authority to AI agents, I believe its board needs to wrestle seriously with three questions.
1. PURPOSE: What problem are we actually trying to solve?
AI should not be the starting point. The business problem should be. Before asking where AI can be deployed, boards and executive teams should be able to answer a much simpler question:
If AI didn't exist, what business problem would we still need to solve?
Is customer churn increasing because response times are too slow? Are employees spending hundreds of hours on repetitive administrative work? Is the organization losing institutional knowledge when experienced employees leave? Are processes preventing the business from scaling? Are leaders making decisions without timely information?
Those are business problems.
AI may be part of the solution, but adopting AI without first identifying the problem creates the risk of acquiring powerful technology and then searching for something for it to do.
This is where the 10X Doctrine begins: Purpose.
Purpose establishes the standard against which organizational decisions are judged. Without it, activity can easily be mistaken for progress.
The same applies to AI. The objective should not be to become an “AI-enabled company.” The objective should be to improve the organization's ability to fulfill its Purpose and those are not necessarily the same thing.
2. PROTECTION: What must AI never be allowed to compromise?
Most conversations about AI risk eventually arrive at data security. They should. But boards need to go further.
The risk isn't limited to a database being compromised. The business itself can be compromised.
Every organization possesses knowledge that contributes to its competitive advantage: pricing logic, customer histories, proprietary processes, contracts, product plans, financial assumptions, source code, operating procedures, strategic plans, internal decision rules, employee expertise, and the accumulated institutional knowledge of how the organization actually works.
Some of that information may never have been formally classified as intellectual property, yet it can still be extraordinarily valuable. So I would ask a board:
What does this business know that we would never willingly give a competitor?
Then follow it with:
Where will that knowledge go when an AI agent is given access to it?
Boards should understand what information an AI system can access, what it retains? Will that information be used to train models? Can the information move outside the organization's controlled environment? And what happens to accumulated agent memory and organizational knowledge if the relationship with a technology provider ends?
This is where AI governance becomes an act of Stewardship.
The 10X Doctrine defines Stewardship as exercising authority with accountability beyond one's own self-interest and with an obligation to leave what is being led in a better state than it was received.
A board therefore has an obligation not simply to pursue the potential value AI can create today, but to protect the assets that allow the organization to create value tomorrow. Competitive advantage surrendered in pursuit of technological advantage is NOT transformation. It is a bad trade.
3. CONTROL: Who owns the intelligence we are creating?
This may eventually become the most consequential of the three questions. As AI agents become embedded in an organization, they may begin accumulating something extremely valuable:
knowledge of how the business operates
Not merely its files. Its workflows. Exceptions. Customer patterns. Decision logic. Institutional memory. Processes. Relationships between information. Potentially even the accumulated lessons from thousands or millions of interactions.
Over time, the organization isn't simply using AI. It is helping create an increasingly sophisticated representation of how the organization thinks and works. Boards should know who ultimately controls that intelligence.
Where does it reside?
Who owns the agent's memory?
Can the provider use what the agent learns?
Can that intelligence be separated from the provider?
Can it be transferred to another platform?
Can it be completely deleted?
What happens if the provider changes its terms, pricing, technology, or business model?
And perhaps most importantly:
Are we building an AI capability that belongs to us, or are we slowly building our business inside someone else's ecosystem?
That is the difference between bringing AI to the business and bringing the business to AI. The 10X Doctrine helps explain why this distinction matters.
Organizational Capacity is:
(Leadership + People + Process) × Technology
Technology does not create Capacity independently. It extends the effect of Leadership, People, and Process. When those elements are sound and technology is properly applied, technology can amplify them. When technology is poorly implemented or misapplied, it can weaken or even invert what the organization has built.
AI may be the most powerful multiplier businesses have encountered in decades. Which means leadership has an obligation to be equally thoughtful about what it is multiplying. That work starts before implementation. It starts with three board-level decisions:
PURPOSE: Why are we doing this?
PROTECTION: What are we unwilling to compromise?
CONTROL: Where will our intelligence live, and who ultimately controls it?
Only after those questions are answered should the conversation move to the next level:
What can an AI agent see?
What can it recommend?
What can it decide?
What can it execute?
And who is accountable when it gets something wrong?
Those are questions of Authority and Accountability, and they deserve their own conversation. But first, the board needs to establish the boundaries within which those decisions will be made.
Your AI advantage does not begin when you buy the technology.
It begins when your board decides how the technology will serve the business without putting at risk what made the business valuable in the first place.
These are not questions to delegate to the IT department. They belong in the boardroom.
At 10X Leadership Consulting, we help boards and executive teams work through these questions before technology decisions become organizational decisions that are difficult to unwind.
If your board hasn't had this conversation yet, it probably should.
