Anthropic secures data center control and Google backing to power its AI infrastructure as it prepares for a $1 trillion（約160兆円） IPO valuation.

What happened: Anthropic has assembled more than a dozen preliminary agreements for U.S. data centers representing more than 1 gigawatt of capacity, and is exploring ways for Google to help support the financial commitments tied to those leases. Google could provide guarantees for Anthropic's lease obligations and co-designs some of the server chips Anthropic may deploy at those facilities.

Why it matters: The infrastructure push reflects Anthropic's need to secure vast computing power as demand for its Claude models surges. The company recently raised $65 billion（約10兆円） in Series H, valuing it at $965 billion（約150兆円） and surpassing OpenAI as the most valuable startup. Anthropic has also confidentially filed paperwork for a U.S. initial public offering, with a potential $1 trillion（約160兆円） valuation that would rank it among the most valuable companies globally.

What to watch: If Anthropic goes public at a $1 trillion（約160兆円） valuation, it could become the second- or third-largest IPO in history, trailing SpaceX and Saudi Aramco. The company disclosed earlier this month that Alphabet was committing up to $40 billion（約6.4兆円） in support.