The competitive landscape for AI startups is shifting. Foundation model releases from companies such as OpenAI and Anthropic are influencing product strategy and fundraising. Founders are increasingly focused on whether their products can maintain value after a platform update.
The core challenge centers on the rapid evolution of foundation models. Capabilities that once differentiated an AI startup can quickly become standard expectations. This necessitates a strategic shift away from solely building the ‘smartest’ model and towards creating lasting value through proprietary data, integrated workflows, and established customer relationships.
This Builders Stage session at TechCrunch Disrupt 2026, featuring Michel Tricot (Airbyte), Rob Toews (Radical Ventures), and Linda Tong (Webflow), examines how to build companies that remain relevant as AI evolves. The discussion will cover identifying defensible AI companies and how to respond when AI giants introduce adjacent market capabilities.
Participants will explore the importance of customer relationships, data ownership, and domain expertise – elements that foundation models struggle to replicate. The session aims to provide founders with insights into building sustainable businesses that aren’t simply features within a larger platform. The focus is on creating lasting value and avoiding the risk of a strong product becoming someone else’s product update.
Key takeaways include understanding where defensibility still exists, how to adapt to evolving customer expectations, and strategies for building companies that remain relevant as AI continues to advance. The session will provide a framework for founders to proactively address this strategic shift.
