Masayoshi Son says firms should track Return on AI as they shift to AI-native operations
SoftBank founder Masayoshi Son used his keynote at SoftBank World 2026 to argue that companies should stop relying on traditional asset-based measures alone and begin using Return on AI, or ROA, as a core management metric. He said businesses need a structured transition toward becoming AI-native enterprises as industries move toward broad AI adoption. In Son’s presentation, Artificial Superintelligence, or ASI, could account for 20% of global GDP by 2040, with annual output reaching 7,000 trillion yen. He also projected a labor transition toward 100 trillion AI agents and 1 billion humanoid robots, describing them as systems capable of autonomous operation, communication, and continuous learning. Son added that supporting this shift would require global AI data center capacity to reach 3 terawatts, with average annual investment of about $5 trillion to push compute to the quetta level, or 10^30. He also said AI can identify thousands of potential vulnerabilities inside systems, which means both individuals and companies need to integrate AI more deeply into daily use and operational design. For companies aiming to become AI-native, Son outlined four steps: maintain a firm AI-led vision, use existing data and industry expertise, automate business tasks with AI agents, and monitor ROA over a three-year period to judge whether spending on AI agents, humanoid robots, infrastructure, and software-hardware systems is producing real productivity and revenue gains.








