SoftBank Group founder and CEO Masayoshi Son has made one of his boldest forecasts yet for the future of artificial intelligence, arguing that global AI investment will eventually reach $5 trillion annually by 2040. Speaking at SoftBank’s annual corporate conference in Tokyo, Son rejected growing concerns that artificial intelligence has entered bubble territory, instead describing current spending as only the early stages of a technological transformation that could reshape the global economy. His remarks reinforce SoftBank’s aggressive strategy of becoming one of the world’s largest investors in AI infrastructure, software, and next-generation computing platforms.

SoftBank Doubles Down on the Long-Term AI Opportunity

Son said annual investment of approximately 800 trillion yen, or roughly $5 trillion, will eventually become economically sustainable as artificial intelligence expands its role across virtually every industry. While he did not provide a detailed methodology behind his projection, he argued that if AI ultimately generates around 20% of global gross domestic product by 2040, the required capital investment would represent only a small fraction of the economic value created.

The comments reflect SoftBank’s increasingly concentrated focus on artificial intelligence after spending the past two years deploying tens of billions of dollars into AI-related businesses. The company has significantly expanded its investment in OpenAI while financing large-scale data center projects and acquiring stakes in robotics companies that could benefit from widespread AI adoption.

Son Rejects Bubble Concerns Despite Massive Capital Spending

Growing investment across the AI industry has prompted increasing debate among investors over whether infrastructure spending is becoming excessive relative to current commercial returns. Major technology companies continue committing record amounts of capital toward advanced chips, cloud infrastructure, and AI model development, raising questions about future profitability.

Son firmly rejected those concerns, arguing that comparisons to previous technology bubbles misunderstand the scale of AI’s long-term impact. He suggested that critics questioning current valuations fail to recognize how deeply artificial intelligence could transform productivity, business operations, and global economic activity over the coming decades.

His confidence follows a career defined by high-conviction technology investments. While SoftBank generated enormous returns through early investments in Alibaba and by introducing Apple’s iPhone to Japan, other investments, including WeWork, resulted in significant losses. Today, the company’s largest strategic commitment centers on OpenAI, where cumulative investments are expected to exceed $60 billion before the end of 2026.

AI Infrastructure and Energy Become the Next Strategic Frontier

Beyond software and computing, Son emphasized that future AI growth will depend heavily on energy infrastructure. He projected that AI data centers could require approximately three terawatts of electricity generation by 2040, equivalent to roughly 1.8 times today’s global electricity consumption.

Initially, natural gas is expected to supply much of that demand, although Son believes nuclear fusion could eventually become the dominant energy source because of its potential for cleaner and lower-cost electricity generation. He also acknowledged emerging concepts such as space-based solar power but argued that practical fusion technology may ultimately provide the more economical long-term solution.

Looking ahead, Son outlined an ambitious vision in which as many as 100 trillion autonomous AI agents operate independently, communicate with one another, and perform complex tasks with minimal human intervention. Whether those forecasts ultimately materialize remains uncertain, but SoftBank’s continued investment strategy signals growing confidence that artificial intelligence will remain one of the defining technological and economic forces shaping global markets over the next two decades.