The Steering Wheel Gap

In my book You to the Power of Two, we describe today’s agentic AI systems as cars with giant engines but no steering wheel and no brakes.

Enormous power. No mechanism to guide it, stop it, or answer for it. Watch what companies do when a mandate lands: They buy the engine. They hire the engine specialists. They benchmark their engine against competitors’ engines. Nobody builds the steering wheel.

The data proves it:

74% of GCC organizations are planning to adopt agentic AI. Only 19% have made it from pilot to full-scale implementation. (IDC research for e& & AWS)

Gartner predicts over 40% of agentic AI projects will be cancelled by the end of 2027 citing escalating costs, unclear business value, and inadequate risk controls. (Gartner, 2025)

Read that list of failure causes again. Model capability isn’t on it. Every failure mode is a leadership failure.

Before you scale a single agent, answer three questions:

1. Accountability— When your agent acts autonomously, who owns the outcome? If the answer is “the vendor” or “the system,” you are not ready.

2. Agency— Does your AI amplify your people’s judgment, or quietly replace it?

3. Adaptability — Can your workforce evolve as fast as your technology stack? The 24-month clock is running. (Dubai Media Office, May 2026)

The engine is not your problem. The steering wheel is.

So here is my question for every leader reading this: Of the three — Accountability, Agency, Adaptability — which one is your organization furthest behind on?