YC just laid out the 7 most powerful moats for AI startup Here’s the founder-friendly version you should have in your head before you ship your next feature: 0) The moat before all moats: SPEED In the beginning you don’t need a moat. You need to move faster than anyone else can. Cursor-style: short sprints, ship daily, learn in public. 1) Process Power The “boring” 10% that takes 90% of the work. Evals, monitoring, recovery flows, compliance, edge cases. If it’s mission-critical and you make it production-grade, it becomes hard to copy. 2) Cornered Resource Something others can’t easily get: proprietary data, relationships, distribution, regulatory access, domain embedding. Not “we fine-tuned a model.” More like “we own the workflow and the labels of what good looks like.” 3) Switching Costs AI will lower old switching costs (agents can migrate data). But it raises new ones when your agent becomes the customer’s operating system (custom logic + integrations + trust built over months). 4) Counter-Positioning Do what incumbents can’t afford to do. Example: per-seat SaaS vs task-based pricing (AI reduces seats, which kills their revenue model). Or second movers winning by focusing on the application layer and shipping a better product. 5) Brand When customers pick you even at parity. “ChatGPT” became the verb. That’s not a feature. That’s compounding advantage. 6) Network Effects (AI-style) Usage → data → better evals/models → better product → more usage. Not “social graph.” More like “feedback loops that competitors can’t replicate without your volume.” 7) Scale Economies Pay big fixed costs once (infra, data, integrations, compliance) and amortize across many customers. The YC punchline: Don’t use “moats” to talk yourself out of starting. Pre-product, you have nothing to defend. Find a hair-on-fire pain, ship fast, then deliberately deepen 1–2 moats.