Inside OpenAI's 10-Year Plan to Build AI Hardware in the US

OpenAI, the company behind ChatGPT, is expanding its ambitions beyond the digital realm and into physical infrastructure.

The firm has launched an extensive request for proposals (RFP) to restructure its hardware supply chain, a move that signals a fundamental shift in how AI companies approach manufacturing and supply chain independence.

The programme aims to create a robust, US-based ecosystem for the components that will power the next generation of AI technology: consumer electronics, robotics and the vast data centres required to support increasingly sophisticated AI models.

Rather than simply being a procurement exercise, the RFP represents a decade-long strategic blueprint designed to localise advanced manufacturing and establish dominance in what OpenAI calls the emerging "Intelligence Age".

Reshaping AI infrastructure

The RFP presents an ambitious vision that extends far beyond software development into the complex realm of silicon, motors and cooling systems.

OpenAI is seeking collaborators which can help them achieve what they characterise as a "Silicon Renaissance" – a return to domestic manufacturing that ensures supply chain stability and national leadership in AI development.

According to the RFP document: "OpenAI has a long-term ambition to establish US-based hardware manufacturing and assembly that reflects US values, supports resilient supply chains and fosters national innovation leadership."

This objective stems from the conviction that, for AI to reach its full potential, the physical infrastructure supporting it must be as advanced as the algorithms themselves.

The initiative represents OpenAI's recognition that hardware independence is crucial for maintaining competitive advantage in the AI sector. By controlling more of its supply chain, the company positions itself to respond more quickly to technological advances and market demands.

The domestic manufacturing focus also addresses growing concerns about supply chain vulnerabilities that have affected technology companies in recent years. OpenAI's approach seeks to mitigate these risks through localised production capabilities.

Critical sectors driving hardware expansion

OpenAI's manufacturing strategy is organised around three essential sectors, each requiring a sophisticated network of suppliers and manufacturers.

In consumer devices, the firm is moving into final assembly, PCB assembly and the creation of advanced displays and optics. This suggests a future where OpenAI-branded hardware becomes commonplace in everyday consumer technology.

For robotics integration, OpenAI is pursuing what it describes as "critical inputs" including actuators, precision bearings, gearboxes and power electronics. These components enable AI systems to operate more effectively in physical workspaces and industrial environments.

Perhaps the most capital-intensive focus area centres on data centre infrastructure – the unseen backbone of AI operations. This encompasses power systems such as generators, transformers and UPS, alongside advanced cooling technologies including chillers and cold plates.

As the RFP document states: "Over the next 10 years, OpenAI seeks to localise significant portions of the manufacturing for its hardware devices and data centres, including key components, modules and final assembly."

This long-term commitment demonstrates the scale of OpenAI's manufacturing ambitions.

The physical reality of AI systems

This initiative follows the trajectory of the Stargate Project, launched in March 2025.

OpenAI has made considerable progress towards its 10-gigawatt power commitment, with capacity plans exceeding the halfway mark by early 2026.

According to the RFP document: "Infrastructure has long been destiny when it comes to America's economic success and that will be especially true in the Intelligence Age."

Through investment in domestic production, OpenAI is bidding to catalyse US manufacturing, modernise the energy grid, create well-paid jobs and strengthen American leadership.

The RFP document emphasises that discussions around AI often begin and end with chips, but the reality is considerably more intricate.

OpenAI notes that advanced AI depends on a much broader ecosystem of physical components: the racks, cabling, networking gear, cooling systems, power systems, power electronics, electromechanical modules and testing and assembly capacity are all required to bring it all online at scale.

Even the most advanced AI models are ultimately constrained by the physical infrastructure that supports them – a consideration reshaping how the industry approaches both development and deployment in the years ahead.