Meta's Muse AI Agent Could Disrupt Aggregator Platforms Like Airbnb, Analyst Says
Financial analyst MBI divested his Airbnb holdings and increased his investment in Meta after testing Meta's Muse AI agent. He observed that Muse can independently navigate third-party sites, bypass aggregator markups by finding direct host bookings at up to 60% savings, and offer hyper-personalized recommendations using cross-platform context. This perspective illustrates how browser-operating AI agents could disintermediate online travel agencies, e-commerce platforms, and food delivery services by reducing user decision friction and avoiding middleman commissions. Additionally, it highlights how Meta could monetize AI agents indirectly without in-app ads by using agent interactions to enrich user profiling and boost overall ad revenues on Facebook and Instagram. During testing, Muse discovered a farm stay that was roughly 60% cheaper when booked directly with the owner compared to Airbnb. However, current technical limitations remain: the browser-based execution process is currently slow, taking up to 14 minutes to compare five hotel options, and early output formats lack rich visual elements like maps and photos.
## BACKGROUND
AI agents are autonomous software tools that leverage modern language models to perform multi-step web tasks, such as filling out forms, querying sites, and placing orders on behalf of users. Meta introduced Muse as a personal AI agent designed for long-running, goal-oriented tasks rather than simple conversational Q&A. Aggregator platforms traditionally earn revenue by taking commissions for matching buyer demand with suppliers, a middleman model that AI agents can bypass by negotiating directly with service providers.