AI is going to drain a lot of moats.
Most companies think their “moat” is safe: data, brand, headcount, processes. Then AI shows up and does in minutes what armies of employees used to do in months. Naval’s point is simple: if a human can do it by following a clear process, AI can probably do it cheaper, faster, and at scale.
The Psychology Behind It
Most founders overvalue what they already built: documents, dashboards, org charts. AI makes these assets commoditized inputs instead of defensible advantages. The real moat shifts to things models cannot easily copy: proprietary distribution, insanely specific customer insight, tight communities, and brand trust earned over years. Build where data is private and emotional, not public and procedural.
Where Moats Quietly Disappear
- Process moats vanish when AI turns complex SOPs into one-click workflows.
- Talent moats shrink as a single person with AI outperforms large average teams.
- Information moats erode when models can learn any public domain in hours.
- Speed moats die because every competitor gets instant execution leverage.
Who Still Keeps a Real Moat?
Apple defends its moat with a closed hardware ecosystem and brand love that AI cannot instantly clone.
Stripe protects its position with deep integrations and hard-won trust in handling money at global scale.