Wharton professor Ethan Mollick observes that companies in the same regulated industry have vastly different AI adoption rates. The key differentiator is whether an executive is willing to assume risk. Without leadership buy-in, IT and legal departments default to blocking new technology.

The primary barrier to AI adoption in large companies is not technological but organizational. Success depends on understanding the 'real' org chart—the informal network of influencers who control data and approve projects, which often differs from the official hierarchy.

AI is a 'hands-on revolution,' not a technological shift like the cloud that can be delegated to an IT department. To lead effectively, executives (including non-technical ones) must personally use AI tools. This direct experience is essential for understanding AI's potential and guiding teams through transformation.

According to Michael Dell, technology for AI transformation is available. The real bottleneck for large enterprises is a lack of leadership courage and a resistant culture. Incumbent processes and incentive structures, like bonuses tied to maintaining the status quo, prevent companies from making necessary bold changes.

Unlike traditional software, AI adoption is not about RFPs and licenses but a fundamental mindset shift. It requires leaders to champion curiosity and experimentation. Treating AI like a standard IT project ignores the necessary changes in workflow and thinking, guaranteeing failure.

Employees hesitate to use new AI tools for fear of looking foolish or getting fired for misuse. Successful adoption depends less on training courses and more on creating a safe environment with clear guardrails that encourages experimentation without penalty.

Relying solely on grassroots employee experimentation with AI is insufficient for transformation. Leadership must provide a top-down motion with resource allocation, budget, and permission for teams to fundamentally change workflows. This dual approach bridges the gap from experimentation to scale.

The most significant hurdle for businesses adopting revenue-driving AI is often internal resistance from senior leaders. Their fear, lack of understanding, or refusal to experiment can hold the entire organization back from crucial innovation.

CEOs who merely issue an "adopt AI" mandate and delegate it down the hierarchy set teams up for failure. Leaders must actively participate in hackathons and create "play space" for experimentation to demystify AI and drive genuine adoption from the top down, avoiding what's called the "delegation trap."