What AI Actually Demands of Leaders

In an extended discussion for AI Frontier Conversations, Microsoft’s Naim Yazbek and Majid Al Futtaim’s Ahmed Galal Ismail make one thing clear: AI exists to solve big problems, not showcase technology. On leadership unlearning, the pilot trap, data as capital, and what it actually means to rebuild an organization in the age of AI.

Naim Yazbek, President of Microsoft Middle East and Africa, sat down with Ahmed Galal Ismail, CEO of Majid Al Futtaim, for a candid conversation on what AI transformation demands of leaders. Five ideas stood out.

1. AI Is a Leadership Problem, Not a Technology One

Organizations where AI sits on the CEO and board agenda are moving. Those that delegated it to technology teams are spending heavily and going nowhere. Yazbek is direct: the difference is not tools or budget – it is where accountability lives. Galal Ismail echoes this from the inside: the re-founding underway at Majid Al Futtaim is as much cultural as technological, shifting from top-down innovation to a model where every leader acts as a founder within their own domain.

2. Data Is Capital. Treat It Like One.

Galal Ismail argues that data, algorithms, and AI agents constitute a new form of intellectual capital that needs to be governed with the same discipline as financial capital — deliberate decisions about where it is deployed and what return it generates. Yazbek makes the behavioral shift concrete: before any performance discussion, his first question is what Copilot reads from the data, not what the PowerPoint says.

3. The Window Is Closing. Fast.

Cloud adoption started in 2006. Real momentum didn’t arrive until 2017. AI has no such runway. Three years in, organizations are already at an inflection point. Yazbek’s view is blunt: get on the train now and learn on the way, or risk irrelevance. Galal Ismail draws the parallel to early e-commerce – companies that waited for perfect ROI before committing often missed the window entirely.

4. Escape the Pilot Trap

Dozens of proof of concept, limited ROI, stalling momentum – the pattern is familiar. The fix is not better measurement, it is fewer, bigger bets. Point AI at the three or four most significant problems in the business, not at the margins.

5. Trust Is Culture, Not Compliance

Both leaders treat trust as an organizational condition, not a governance checkbox. Galal Ismail makes the point sharply: in an AI-enabled world, individual errors cascade at scale. Good intentions are not a safeguard – culture is. Yazbek adds the institutional layer: Microsoft’s commitment is that customer data improves how that customer uses the model, not the model itself. Transparency about mistakes, both agree, is what separates organizations that hold credibility from those that lose it.

The Hardest Skill: Unlearning

Both leaders candidly shared what they have personally had to unlearn. Yazbek: stop allowing decisions without data; bet on potential over readiness; use Copilot to plan the day before instinct kicks in. Galal Ismail: stop needing to be in the room. Absence, he has found, creates accountability rather than a vacuum – and gives teams the space to make better decisions than they would have otherwise.

AI Frontier Conversations brings together leaders shaping the region’s technology and business landscape.