Why it matters now
When AI sits with IT or with one enthusiastic person, it tends to produce tools rather than change. The decisions that matter, such as which problems to solve, what to stop, how roles change and how much risk is acceptable, belong to the people who run the business. The companies I see making progress are the ones where the top team treats AI as part of running the company, not as a side project.
- 72% of CEOs say they are now the main decision-maker on AI in their company, double the share of the year before.BCG, AI Radar 2026 ↗
- At larger companies, CEO oversight of AI governance was the factor most strongly linked to EBIT impact from AI.McKinsey, The State of AI: How organizations are rewiring to capture value, 2025 ↗
- High-performing companies are three times more likely to say their senior leaders show ownership of AI.McKinsey, The State of AI in 2025 ↗
Common mistakes
- Handing AI to IT or to a single “AI person” and expecting the business to change.
- Naming an owner with no budget and no authority, so decisions never stick.
- Letting every unit run its own tools and pilots with no shared standards.
- Leaving it unclear whether the business or the technology team answers for results.
What to do instead
- Keep the agenda with the top team. Make AI a standing item for the executive team. Strategy, priorities and acceptable risk are decided there.
- Give business leaders the outcomes. Each business leader owns the results AI should deliver in their area, in the same way they own their other targets.
- Build a small central hub. A central team owns shared platforms, data, standards, vendor choices and risk, and helps the business units move faster.
- Name one accountable executive. Give one person a clear mandate, a budget and a regular review with the CEO. This is about accountability, not a title.
- Write down decision rights. Record who decides on priorities, spending, vendors and risk, so nobody has to guess.
Your first step this week
Write down who currently decides on AI priorities, budget and risk in your company. If the answers are vague or point to different people, bring that page to your next executive meeting.
Related guides
- Should we hire a Chief AI Officer?
- What is my own role as CEO, and is my leadership team AI-literate enough?
- How do we govern AI risk and use AI responsibly?
All AI leadership principles and guides · 100 AI use cases by function