Mastering AI in strategy: Not what, but how

By Carolyn Dewar, McKinsey & Company senior partner and co-author of The New York Times bestseller, CEO Excellence: The Six Mindsets That Distinguish the Best Leaders from the Rest.

I get it. AI fatigue is real. If you’re a CEO, you’ve probably sat through more breathless AI presentations than you can count—half of them promising the singularity, the other half a glorified chatbot. But buried beneath the hype are real, and perhaps counterintuitive, ways in which AI is reshaping strategy.

Let’s just agree—AI is not going to replace human judgment in the complex craft of corporate strategy. But there’s no doubt it?is?becoming a powerful tool for leaders who understand how to wield it effectively. The smartest CEOs aren’t just asking?what?AI can do; they’re figuring out?how?to use it to rethink strategy development in ways that actually matter.?

For CEOs wanting to understand how AI is transforming strategy development, the biggest mindset shift is that AI isn’t just about enabling faster or more insightful decision-making—it’s about changing?how?we make decisions, and then, how we move on them. A few key insights stand out:

?1. AI is making human judgment more important, not less

There’s a common fear that AI will take the “art” out of leadership, replacing human judgment, logic, and experience with data-driven certainty. But the reality is the opposite. AI can analyze vast amounts of data and identify patterns humans might miss. It can collect data, find patterns, and generate insights quickly, freeing up strategists to hone their decision-making processes. But it can’t replace the bold commitment required from CEOs when choosing to double down or pivot.

As access to data becomes increasingly democratized and AI adoption spreads, proprietary data insights, best-in-class strategic thinking and human judgment will be the critical points of distinction for leaders. The best CEOs are using AI to inform their decisions and help define bold moves they may otherwise not have considered—it won’t replace instinct, or strategic courage.

?2. AI in strategy isn’t only about prediction—it excels in experimentation

?A lot of AI conversations focus on prediction: Can it forecast market shifts? Customer demand? Competitive moves? While AI can certainly help refine predictions, in the world of strategy its true power isn’t limited to looking into the future—it’s an invaluable tool for?testing?different possible futures.

In 2009, a major beverage company bridged the gap between testing what’s possible and making it happen. Moving beyond the use of AI solely for forecasting demand, they deployed it to run thousands of rapid product experiments. In one such experiment their customers became the subjects, with AI-powered smart vending machines that allowed users to customize flavors. Data from these interactive transactions was fed into their R&D pipeline—directly informing product strategy from AI-driven, real-world experimentation, in real-time. Fast-forward to 2025 and this experiment has its own app and some 50,000 vending machines globally.

The real question addressed here isn’t?what will happen??but?what should we try next? AI can be used to help companies run more strategic experiments, faster and cheaper than ever before—opening a world of possibilities for creative, forward-thinking moves.

The best strategists aren’t simply consuming AI-driven insights—they’re actively shaping the questions AI is answering and being definitive about the problem the business wants solved.

?3. AI can expose blind spots, but only if you ask the right questions

?If you’re not getting useful answers from AI, the problem might not be the technology—it might be that you’re not asking the right questions. AI is brilliant at spotting patterns, but it won’t magically surface hidden risks or challenge entrenched assumptions, unless you design it to do so.

When large language models first launched, it quickly became clear that the machine can sometimes?hallucinate—producing confident-sounding but incorrect answers. Some companies assumed that simply improving model accuracy would solve the problem. Others took a different approach.

One global law firm started using AI for contract review but found that the technology frequently made mistakes in complex legal language. Instead of abandoning it, they refined their approach. Rather than relying on AI to?answer?legal questions outright, they trained it to act as a second set of eyes—flagging inconsistencies, comparing similar contracts, and suggesting clauses for further human review. The key shift? They stopped treating AI as an?omniscient decision-maker?and started using it as an?intelligent challenger?to surface blind spots.

The best strategists aren’t simply consuming AI-driven insights—they’re actively shaping the questions AI is answering and being definitive about the problem the business wants solved.

4. The real advantage isn’t AI itself, it’s how fast you act on what it finds

The companies seeing the biggest returns from AI aren’t necessarily the ones with the best models. They’re the ones that have rewired their organizations to act on AI-driven insights faster than the competition.

One fast-fashion giant saw an opportunity to use AI for more than predicting trends; they built an entire system around responding to AI insights at lightning speed. Like many competitors, their AI models analyze sales data, customer preferences, and social media trends daily. But what really sets them apart is their ability to act, with speed.

With a supply chain structured to take AI-driven insights—like a spike in demand for a particular style—they can move from design to in-store shelves in as little as two weeks. Compare that to traditional retailers that take months to roll out new products. The result? The company isn’t just using AI to understand what customers want; it’s?out-executing?competitors by acting on those insights faster than anyone else.

?AI alone won’t make you faster. How?you use it will.

Like all tools, AI is only as good as their users

AI won’t make you a great strategist—but when used well, it can sharpen your edge. The leaders who thrive in this new era will be those who understand how the technology can significantly augment their decision prowess and create space for breakthrough ideas. The generic application of AI tools can only lead to generic strategy. Mastery requires understanding how AI works and then applying it to strategy development in a way that is specific to what the organization needs.

This includes growing AI know-how within the strategy team and developing a proprietary insight ecosystem that brings a competitive edge as the volume and accessibility of data continue to intensify—then harnessing the insights to be more creative around bolder, bigger, faster bets.

?If you’re tired of the hype but don’t want to get left behind, this article is worth a read and you can also listen to the authors discussing how AI is transforming strategy in this podcast.

?