SoftBank Launches Roze AI: Robots for Data Center Construction and $100 Billion Valuation

SoftBank is planning to create a new company called Roze AI in the United States, which will automate data center construction using autonomous robots. According to Financial Times and The Wall Street Journal, the group is discussing not only launching the project but also its potential IPO in the second half of 2026 with a target valuation of around $100 billion.

Why SoftBank Needs Roze AI

The idea looks almost like a closed loop in the AI boom: to build artificial intelligence systems, you need data centers, and to build data centers faster, SoftBank wants to use artificial intelligence and robots. The task is practical: to accelerate the deployment of server infrastructure in the United States, where demand for computing power is growing faster than new facilities are being created. For SoftBank, this is a logical continuation of a strategy in which AI is not just models and chips, but the entire physical foundation around them.

Whoever controls the speed of data center construction, connection, and launch gets an advantage across the entire chain. In such a market, success comes not only from the best algorithm but also from the ability to put a new facility into operation faster than everyone else.

What Is Known About the Project

According to Financial Times, the new entity will be called Roze AI. The Wall Street Journal reports that its goal is to make American data center construction more efficient, in part through autonomous robots that will help build server campuses. For now, this is one of the few cases where a major technology group is trying to turn the process of building AI infrastructure itself into a separate product and an independent public company. Here is what is currently known about the project:

• The company is being created for the US market
• Focus — data center construction and deployment
• One of the key tools — autonomous robots
• Some executives want to take the business public in the second half of 2026
• The desired valuation could be around $100 billion

However, key details are still missing. It's unclear which specific robotic systems will be used, whether Roze AI will build its own automation stack or integrate contractor solutions, and what share of work can realistically be removed from the manual process. But the problem statement itself is telling: SoftBank wants to earn not just from AI services or chips, but from a machine that helps physically launch the infrastructure for them.

Where the Weak Points Are

The project's ambition is huge, but so are the questions about it. Financial Times notes that within SoftBank there is skepticism about both the valuation and the timeline for the IPO. The company is still being formed, and yet the conversation is already about a valuation comparable to the largest public players in the infrastructure market. For such a valuation, companies typically don't wait for a concept but for clear revenue, confirmed demand, and a sustainable business model.

"Some within

SoftBank are skeptical about the valuation and the proposed IPO timeline."

Doubts are reinforced by SoftBank's own track record, which has made very bold bets more than once. The company has had major successes, but also expensive failures: the publication recalls that the group once invested hundreds of millions of dollars in Zume — a startup about AI pizza delivery that closed in 2023. This doesn't mean Roze AI will meet the same fate, but it shows the gap between a grand visionary idea and actual execution.

There is also a purely industry-level risk: data centers are built not just with concrete and robots. You need land, permits, power supply, network connections, transformers, server equipment, and long coordination with contractors. Even very strong automation won't eliminate all bottlenecks.

But if SoftBank's new business can reduce launch timelines by even months, demand for it from major cloud operators and AI labs could be very high.

What This Means

The AI market is increasingly moving into hardware and physical infrastructure. After the race for models and chips comes the race for the speed of building capacity. If SoftBank can package this process into a separate business, one of the most valuable AI assets could be not another chatbot, but a company that builds buildings for everyone else.