Strava Restricts Data Access to Combat AI Scraping Ahead of IPO

Fitness tracking platform Strava is implementing significant changes to its data access policies, moving to lock down its website and API to prevent unauthorized AI scraping. The company is shifting public-facing data, such as user profiles and club listings, behind an authentication wall. Additionally, Strava is transitioning its developer ecosystem to a paid model, introducing a flat monthly fee of $11.99 for API access while simultaneously retiring specific endpoints to better control how third-party applications interact with its infrastructure.

CEO Michael Martin has framed these measures as a necessary defense against the "ruthless" data harvesting practices of AI firms, which he argues have degraded site performance and ignored established terms of service. By citing instances of aggressive scraping—including specific allegations against AI search startup Perplexity—Strava is positioning itself as a guardian of user privacy and system integrity. The company also plans to adopt the Model Context Protocol (MCP) to provide a more structured, controlled way for AI assistants to access data, moving away from the chaotic, high-volume scraping that has strained its servers.

These policy shifts come as Strava prepares for a potential IPO, suggesting a strategic effort to demonstrate data discipline and long-term value to prospective investors. By tightening control over its proprietary data, Strava is following a growing trend of digital platforms prioritizing the monetization and protection of their assets against AI model training. While the move risks alienating some segments of its developer community, the company is attempting to balance these restrictions with a 90-day grace period and a flat-fee structure that it hopes will remain accessible compared to the more prohibitive pricing models seen elsewhere in the industry.