Igor Babuschkin, founder of the artificial intelligence startup River AI and former lead at Elon Musk’s xAI, is seeking to reshape the way A.I. technologies are developed and controlled. Based in Palo Alto, California, River AI aims to democratize access to advanced A.I. by creating open-source software and hardware solutions that individuals and businesses can operate independently, without reliance on centralized corporate oversight.

With backing of approximately $1 billion from venture capital firms and technology partners, including chip maker Nvidia, River AI is building a new type of server designed for use in homes and small businesses. This system would allow users to run sophisticated A.I. models locally, maintaining control over their private data and the way the technology functions. Babuschkin envisions consumers interacting with these systems similarly to existing digital assistants like Apple’s Siri, but with the added ability to customize or “retrain” A.I. according to personal preferences.

Babuschkin, 35, stressed the importance of returning control over A.I. systems to the end user. He noted that the vast personal data processed by these technologies calls for enhanced privacy and user authority. The open-source approach, he said, would enable individuals to define how A.I. behaves rather than accepting restrictions set by major industry players such as OpenAI and Anthropic, whose products typically remain under strict corporate control.

River AI currently employs about 20 staff members, many of whom previously worked at prominent A.I. firms including xAI, OpenAI, and Tesla. The company is several months away from launching its open-source offerings but aims to develop systems capable of adapting behavior dynamically as users engage with them, potentially even responding to spoken commands.

The push for open-source A.I. has gained momentum recently, partly in response to Chinese startups releasing powerful open-source models that challenge U.S. dominance in the field. While several American companies have released open-source tools in the past, critics argue that U.S. firms have not kept pace with these developments, potentially slowing innovation and limiting access to cutting-edge A.I. technologies outside major corporations.

Advocates of this decentralized model argue that the prevailing approach by companies such as OpenAI and Anthropic—restricting access due to safety concerns—could concentrate excessive power within a few firms and stunt broader industry growth. They contend that open collaboration fosters faster innovation and provides essential tools for businesses facing cybersecurity threats.

Hemant Taneja, CEO of venture capital firm General Catalyst, which leads River AI’s funding round, described Babuschkin as uniquely qualified to deliver a competitive American alternative to Chinese open-source offerings. “Hopefully, this is America’s answer,” he said.

Babuschkin has also worked as a researcher at OpenAI and Google. He expressed skepticism about the current trajectory of large A.I. companies, emphasizing public concern over handing control of powerful technologies to a small number of organizations. “We don’t want these A.I. companies to rule the world and control this superpowerful technology,” he said, underscoring River AI’s mission to create a more accessible, user-centric A.I. ecosystem.