How Old Leadership Haunts AI Success

The AI graveyard is littered with promising projects that died in the proof-of-concept stage. We blame AI technology and its lack of reliability, but a more dangerous, human-centric problem is emerging as the real bottleneck.

For the last 18 months, I've been on the front lines of AI transformation with organizations across a variety of industries. While most leaders grasp that AI changes everything, precious few have changed how they lead.

The End of the Answer Economy

For decades, leaders were promoted because they had the answers. They knew the market, the product, the numbers. Seniority meant knowing more than the people around you. AI breaks that model in a fundamental way.

When anyone on your team can get an answer from ChatGPT in thirty seconds, having the answer is no longer your competitive advantage. Framing the right question is. The quality of every AI output is determined by the quality of the human input, and the most important input isn't data. It's the question.

This principle is central to my work with Boston Consulting Group (BCG), coaching organizations on AI adoption and digital growth. Their methodology is built around the SCQ (Situation-Complication-Question) framework, which demands a relentless focus on framing the problem correctly before attempting a solution. It’s a disciplined approach that stands in stark contrast to the answer-first culture I see everywhere else.

But in company after company, I still see leadership meetings dominated by answer-giving. Senior people asserting what they know and demonstrating expertise. The room rewards the person with the most confident answer, not the person who reframes the problem entirely.

Organisations automate the wrong things. They apply AI to processes that feel familiar rather than questioning whether those processes should exist at all. They get efficiency gains on yesterday's problems while missing the strategic questions that would unlock tomorrow's growth.

The leader who asks "What problem should we actually be solving?" will outperform the one who says "Here's how to solve it faster."

The Silo Tax gets more Expensive

Most senior leaders earned their roles through deep expertise in a single domain. That depth still matters, but AI is quickly commoditising domain knowledge. What remains scarce is the ability to connect across domains, and that’s where the highest-value opportunities lie.

For example, customer churn may link to supply chain delays, which tie back to pricing decisions, themselves driven by policies set in another department. No single function sees the full picture. Only leaders who think across boundaries can act on it. Organizations should accelerate cross-functional assignments to build this capability in their leaders.

Each function sub-optimises for its own metrics, leading to local gains but missed system-wide opportunities. At the same time, AI investments become siloed, teams build their own tools, models, and data pipelines, so the organisation pays for fragmentation instead of integration.

I’ve seen this across companies of all sizes: long lists of disconnected AI initiatives tracked in Excel or Google Sheets, with little to no cross-functional impact. In many cases, initiatives are even allocated based on fairness across departments rather than overall value or ease of implementation.

Leaders, Integration Architects ,don't need to be the deepest expert. They need to be the person who sees how marketing, product, operations, and data connect, and who has the authority and the instinct to make decisions that optimise for the whole.

Nobody Designed the Handoff

Leaders make decisions and that is their job, but AI introduces a new layer of complexity: decisions are increasingly being informed, or recommended, by algorithms. And most organisations have not designed the process for how human judgment and machine intelligence work together.

The questions that nobody is answering: When does AI recommend and a human approves? When does AI act autonomously? When does a human override the algorithm? And who is accountable when an AI-informed decision goes wrong?

Chaos dressed up as innovation. The AI recommends something the salesperson disagrees with, and there's no process for resolving the tension. So the salesperson ignores the AI, the product team gets frustrated, and the investment generates no return. Not because the technology was bad, but because nobody designed the decision architecture around it.

The leader's job is no longer just to make the call. It's to design how calls get made when humans and machines are both in the loop. This is an entirely new leadership competency, and almost nobody is building it deliberately.

You announced AI but your People heard something else.

Every transformation has a communications strategy. But there is a vast difference between announcing change and making meaning from change. And AI transformation triggers something deeper than most change initiatives, because it touches people's sense of professional identity.

When you tell your team "we're implementing AI," what most people hear is not "exciting new tools." What they hear is: Am I going to be replaced? Is my expertise still valued? Will my job exist in two years? No amount of town halls or FAQ documents will address that fear, because it's not an information problem. It's a meaning problem.

The gap between the announcement and the meaning is where trust erodes. It's in that gap that rumours start. It's where your best people quietly update their LinkedIn profiles. It's where the 60% of your organisation who are willing but waiting, decide to wait a little longer. And every week they wait, the transformation stalls further.

Meaning-making requires three things that announcements don't: specificity: "here's what's changing in our team"; honesty: "here's what I don't know yet"; and personal commitment: "here's what I'm going to do to support you through this." That takes individual conversations, not mass communications. And it takes courage, because you're admitting uncertainty while simultaneously asking people to move forward.

The Pattern I Keep Seeing

Technology teams build impressive AI capabilities. Leadership announces the transformation with enthusiasm. And then nothing fundamentally changes in how decisions get made, how functions collaborate, how people experience the change, or what questions the organisation is asking. The AI gets deployed. The leadership doesn't evolve. And twelve months later, the company has AI tools and old-model leaders, and the gap between them is where value goes to die.

The organisations that are getting this right, and there are a few, share a common trait. Their leaders treat AI transformation not as a technology initiative they oversee but as a leadership transformation they undergo. They're changing themselves, not just their tools.