With nearly all new products incorporating AI, startups pitching AI as their main edge overlook the fundamental investor demand: moats that withstand well-funded competitors. Analysis of over 570 AI-focused startups reveals only two effective moats today—counter-positioning and network economies—while conventional advantages like proprietary data and switching costs often fall short.
- Counter-positioning creates moats by making incumbents’ replication economically damaging.
- Network economies generate winner-take-all dynamics by increasing value with participants.
- Proprietary data and switching costs rarely guarantee durable moats in the AI era.
What happened
Analysis of 576 venture-backed AI B2B startups raising $50 million or more since 2025 reveals the erosion of AI as a unique competitive advantage. Nearly all new products integrate AI, making it a baseline rather than a moat. Investors instead prioritize startups that can build defenses difficult for deep-pocketed rivals to replicate.
By applying Hamilton Helmer’s 7 Powers framework and insights from a large product community, researchers identified only two types of moats reliably protecting startups today: counter-positioning and network economies. These moats leverage business model innovation and network effects over proprietary technology.
Why it matters
Counter-positioning power arises when a startup’s model forces incumbents to choose between matching the innovation and undermining their existing lucrative lines, creating a rational barrier to competition. Examples include AI-native insurers selling directly to employers or AI-driven revenue management systems that would cannibalize consulting fees if copied by traditional vendors.
Network economies emphasize B2B platforms linking companies, where each added participant increases the network’s value for all others. This creates a potent compounding advantage difficult to challenge. Securing early commitment from both sides of the network remains challenging but is key to establishing a durable moat that transcends simple data or technology advantages.
What to watch next
Founders should evaluate whether a large competitor could replicate their model without incurring unsustainable costs or economic harm, and prioritize building networks that amplify value with scale. Those relying primarily on proprietary data or customer switching costs must invest in compounding advantages that AI cannot easily erode.
Investors and founders alike will continue scrutinizing moats beyond AI integration, focusing on business model innovation and network effects as key long-term defenses. The success of startups embracing these moats will influence venture trends and the evolution of competitive dynamics in AI-powered industries.