SAN FRANCISCO — Igor Babushkin, a co-founder of xAI, has launched River AI with $1.1 billion in initial funding to build personal AI systems that users own and control.
The company plans to develop AI agents that learn continuously from individual users, operating according to user interests rather than serving the priorities of large tech platforms.
Babushkin wants to place a new type of computer inside people's homes, allowing individuals to run AI locally without oversight from River AI or other large entities.
The startup's team includes former staffers from xAI and Tesla Inc. bringing experience from high-profile AI and technology development environments.
River AI's core model differs from downloading a standard application. Users can retrain their AI agents directly, creating a digital assistant that adapts to individual preferences over time rather than reflecting platform-level defaults.
The $1.1 billion raise positions River AI against existing AI platforms where user data and agent behavior are controlled by the provider. The company is betting that ownership structure, not just performance, becomes a meaningful competitive differentiator.
Building a personal AI stack that learns continuously and runs locally presents real technical challenges: managing hardware requirements, optimizing on-device model training and protecting user data without centralized infrastructure.
If the model works, user-owned AI could support monetization built around personalized services rather than aggregated data — a direct challenge to the economics of current large-scale AI platforms.
