A Message to CEOs on AI Adoption

Image credit: OpenAI

McKinsey published a report [1] around mid June this year on “A CEO playbook to solve the gen AI paradox” and why fewer than 10% of vertical AI use cases make it past the pilot stage for Fortune-500. The report is solid, and I agree with much of it — especially around the “human bottleneck”.

But here’s what I think, things work a little differently for small to mid-sized companies (say, a few ten to few hundred engineers):

In a mid-size org, that human problem looks a bit different. With leadership closely involved in day-to-day decisions, the bottleneck often isn’t governance or scale — it’s executive mindset and cultural design. When leaders default to “what worked before,” potential technical ideas can be deprioritized not because they’re unfit, but because of cognitive biases and a low tolerance for visible failure.

Two patterns we often see:

• Solo experiments & hidden failure: engineers run isolated pilots — when something fails, it feels personal. People stop experimenting.
• Beliefs and perceptions of the C-suite: If leadership anchors on what worked in the past, great ideas from engineers may get ignored — not for lack of merit, but because of bias.

So what does success look like ?