Elon Musk’s Former AI Co-Founder Raises $1.1bn To Break Big AI’s Grip

Former xAI Mastermind Secures $1.1bn To Build AI You Actually Own

The $1.1bn AI Bet That Quietly Validates Elon Musk’s Vision

Igor Babuschkin, one of the original architects of Elon Musk’s xAI, has secured $1.1 billion for a new company attempting to overturn the way artificial intelligence is built and controlled. His startup, River AI, wants customers to train, customise and ultimately own models shaped around their own information.

The scale of the investment is remarkable, but it also delivers an important verdict on Musk’s judgement. One of the engineers he recruited to build xAI has emerged with backing from some of the world’s most influential technology investors, showing that Musk did not merely launch another chatbot—he assembled an extraordinary concentration of valuable AI talent.

River AI Wants To End The Rental Model

Most people and businesses currently access advanced AI on somebody else’s terms. They send information to a centrally controlled model, pay for access and remain dependent on the company operating the system.

River AI is betting that this structure will not last. Babuschkin argues that future models should learn from their individual users, understand their preferences and remain under their control rather than functioning as identical assistants supplied to millions of customers.

Its first product is an application programming interface designed to help developers and businesses adapt open-weight models around their own work. River claims customers can conduct complex reinforcement-learning training runs in 15 to 20 minutes without maintaining a dedicated infrastructure team.

The company also claims its system can be between two and four times more cost-effective than closed alternatives. Those performance and cost comparisons remain company claims, but the investors backing River have clearly decided that the underlying opportunity is substantial.

Nvidia And AMD Are Backing The Vision

The funding was led by General Catalyst and AMP PBC. Nvidia, AMD Ventures, Y Combinator and Singaporean investment company Temasek also participated.

That investor list is particularly significant. Nvidia and AMD occupy central positions in the global market for the processors required to train and operate powerful AI systems, giving both companies an unusually close view of where demand may be heading.

Their participation does not guarantee River’s success, but it gives the company access to capital and relationships extending across the AI infrastructure industry. River has declined to disclose the valuation attached to the round.

The New AI Would Learn From Its Owner

Babuschkin’s vision extends beyond allowing companies to fine-tune existing models. River ultimately wants to create personal agents that continually learn how an individual works, remember that person’s goals and complete complicated assignments without constant supervision.

The company describes this as intelligence that belongs to the user rather than being rented by the token. That distinction could become increasingly important as AI systems gain access to private messages, business records, personal preferences and sensitive professional information.

River believes personal AI may eventually require changes across the entire technology stack, including training systems, models, consumer products and hardware capable of keeping intelligence closer to the person using it. It is an ambitious attempt to redesign the relationship between humans and machine intelligence rather than simply release another general-purpose chatbot.

Musk’s Influence Is Still Visible

River is an independent company, and Babuschkin had already worked at Google DeepMind and OpenAI before helping Musk establish xAI. It would therefore be inaccurate to attribute his entire career or River’s technology to Musk.

Yet the philosophical overlap is difficult to ignore. xAI released the weights and architecture of its original Grok-1 model under an open licence, challenging the increasingly closed direction taken by several leading AI developers.

River is carrying that argument further. Instead of merely providing access to an open model, it wants users to adapt, control and potentially operate intelligence developed specifically for them.

Babuschkin also gained experience building at extreme speed inside xAI. Musk’s company constructed its own training systems and entered the frontier-model contest despite beginning years behind its largest competitors.

The fact that a member of that original team can now attract $1.1 billion is not simply a story about an employee leaving. It is evidence that Musk created an organisation whose technical alumni are considered capable of building major AI companies of their own.

A Different Front In The AI War

River is not trying to defeat xAI by producing a direct Grok replacement. Its immediate opportunity sits at a different layer: giving developers and enterprises the tools to transform open models into specialised systems.

If that approach works, the AI market could become less dependent on a handful of central laboratories. Businesses might combine powerful open models with their own information, training and safeguards instead of sending every task through the same externally controlled services.

There are still serious risks. River must prove that its platform can deliver its claimed advantages at scale, protect sensitive training data and convert an enormous investment into a commercially sustainable business.

Personalisation also creates difficult questions about privacy and security. An agent capable of learning a person’s habits and acting independently could be extremely useful, but the same depth of knowledge would make failures, breaches or manipulation far more consequential.

Musk’s Talent Bet Keeps Paying Off

Elon Musk has repeatedly argued that artificial intelligence is too important to be controlled by a small and unaccountable group. River’s user-owned approach is not identical to xAI’s mission, but it pushes against the same concentration of technological power.

The funding also reveals something broader about Musk’s companies. Their value is not limited to the products they release; they attract engineers willing to pursue exceptionally difficult ideas at extraordinary speed.

Babuschkin is now taking that experience into a company of his own. If River succeeds, it will not diminish what Musk built at xAI—it will demonstrate how influential the team he assembled has become across the next generation of artificial intelligence.