AI Should Reshape the CEO Role—But Not in the Way You Think

Data used to live in the warehouse. Then it moved into dashboards. Now it’s creeping into the boardroom.

Are we entering the era of the “Algorithmic CEO”? Should CEOs fear for their jobs? Or will they be the “human in the loop” when it comes to decision-making? It’s not a robot replacing leadership — but AI systems quietly shaping which strategic initiatives or directions get funded, which markets get entered, and which bets get killed.

For years, executives have struggled to make data-based, unbiased decisions regarding strategic growth options. Traditional financial metrics paint the picture of what worked. They don’t tell you what’s going to work. So, justifiably without an innovation accounting system in place, leaders default to instinct, politics, or whichever initiative screams the loudest in the room.

AI promises to fix that. It promises to evaluate early signals continuously. To model adoption curves. To stress-test downside exposure. And to compare growth bets across the entire portfolio — all while staying unbiased and unmoved by fancy PowerPoint decks and office politics.

Sounds rational. Efficient. Safer.

But here’s the twist.

Breakthrough, early-stage initiatives (innovation) look terrible on paper in terms of data. They are small. They are inefficient. And they always under-perform relative to the existing core business.

So, if you train an algorithm to help in options planning and decision-making, you are probably going to train it on historical performance data. Most, if not all, LLMs are trained this way. In that case, the CEO’s co-pilot AI will favor incremental improvements over disruptive bets — every day of the week.

However, growing beyond the core and defining the company’s next S-curve requires breaking the pattern of the incumbent business model.

So the real danger isn’t that AI will replace CEOs. The danger is that, with AI in the loop, CEOs will become overly rational — and even more reluctant to invest beyond the core business.

And overly rational companies rarely reinvent industries.

Therefore, we shouldn’t be discussing whether it’s better to have AI augment decision-making or rely on intuition. Instead, we should be discussing how we design decision authority in the AI world and what role AI should play in decision-making if we want our companies to stay relevant in the future.

Here are three things to consider when you bring AI into the boardroom with the hope of improving your decision-making process and helping your company define its next strategic moves:

1. Use AI to rank assumptions, not ideas. Rather than letting AI judge which ideas are “good” or “bad,” use it to evaluate the assumptions underlying each initiative. For example, if a new product idea depends on user adoption doubling in six months, the AI can analyze historical adoption trends, market signals, and competitor data to highlight which assumptions are weak, strong, or uncertain. This approach ensures that the organization retains knowledge about why decisions were made, not just which ideas were chosen. Over time, the company builds a repository of validated and invalidated assumptions, helping future leaders make faster, smarter decisions without reinventing the wheel.

2. Separate activity metrics from impact metrics. AI can track thousands of operational or activity metrics, but the real power lies in connecting those activities to actual business impact: revenue growth, customer retention, or market share expansion. By distinguishing activity from impact, AI retains institutional knowledge about what truly moves the needle. Leaders can revisit this knowledge in future decisions, ensuring that lessons from past initiatives aren’t lost in the noise of busy dashboards or vanity metrics. Over time, this helps the company remember which levers consistently drive success, even as teams and strategies change.

3. Let algorithms inform portfolio balance — but reserve human judgment for asymmetric bets. AI excels at analyzing patterns and optimizing for incremental improvements. It can recommend portfolio adjustments that maximize expected returns based on historical data, ensuring knowledge about past decisions, performance trends, and risk exposure is preserved and leveraged. However, truly transformative, asymmetric bets — entering new markets, developing disruptive products, or reshaping business models — require human judgment informed by experience, intuition, and context. By keeping humans in the loop for these high-stakes decisions, the organization retains the nuanced knowledge that AI can’t quantify, preserving strategic wisdom that might otherwise be lost to pure data-driven optimization.