What does Satya Nadella mean by "token capital"?

Microsoft CEO Satya Nadella defines token capital as the proprietary AI systems, models and capabilities a company develops and owns. He argues that these AI assets will become as important to businesses as traditional forms of intellectual property and expertise.

Why does Satya Nadella believe companies need token capital?

Nadella believes companies that build their own AI capabilities will be better positioned to create long-term value. Rather than relying entirely on external AI providers, firms can develop systems that capture and improve their own knowledge over time.

How is token capital different from human capital?

According to Nadella, human capital consists of employee knowledge, judgment, relationships and experience. Token capital refers to the AI systems and digital intelligence a company owns, with the two working together rather than replacing one another.

Does Nadella think AI will replace human workers?

No. Nadella specifically argued that human capital becomes more valuable as token capital grows. He believes human judgment and direction remain essential because AI systems need people to guide decisions, strategy and learning.

What is a "learning loop" in Nadella's AI strategy?

A learning loop is a system where human expertise and AI continuously improve each other over time. Companies can use proprietary data, workflows and feedback to create AI systems that become more useful and accurate with every interaction.

Image for illustrative purposes only. Not a real photo.

Microsoft CEO says firms need AI token capital

Written by Liezl GambePublished Jun 15, 2026, 7:00 AM

Microsoft Chief Executive Officer Satya Nadella has urged companies to build what he calls "token capital," arguing that proprietary AI capabilities will become a core asset alongside human expertise in the emerging AI economy.

Nadella described token capital as the AI systems, models and technological capabilities a company develops and owns, while human capital consists of employee knowledge, relationships and judgment accumulated over time.

“Importantly, human capital does not become less valuable as token capital grows… Without human direction, you have compute running in circles,”

Said Microsoft CEO, Satya Nadella.

The executive argued that businesses should focus less on selecting the most powerful foundation model and more on creating learning systems where human expertise and AI capabilities continuously reinforce one another.

“You can offload a task, or even a job, but you can never offload your learning. The future of the firm is the ability to compound that learning across people and AI,”

Nadella added.

He proposed that organisations build agentic AI systems trained on proprietary data and workflows, allowing them to retain control of intellectual property and institutional knowledge while benefiting from advances in AI models.

Nadella also warned against a future where most economic value is concentrated among a small number of AI model providers, arguing that companies must maintain ownership of their learning loops to preserve competitiveness as enterprise AI spending continues to accelerate.