The next workforce could include millions of AI agents, says Microsoft's Satya Nadella

As businesses race to integrate artificial intelligence into daily operations, Microsoft CEO Satya Nadella believes the biggest challenge may not be building AI systems but managing them.

Speaking on the Possible Podcast with LinkedIn co-founder Reid Hoffman, Nadella painted a picture of a future workplace where AI agents vastly outnumber human employees. According to him, organisations may soon find themselves overseeing not thousands but potentially millions of AI-powered assistants operating across departments, workflows and business functions.

The Microsoft chief suggested that companies will need entirely new approaches to governance as AI becomes embedded in the structure of modern enterprises. Rather than viewing AI as just another software tool, Nadella argued that organisations should start preparing for a world in which digital workers play an active role in getting work done.

More from Tech

Why AI agents will need oversight

One of Nadella's central arguments was that AI agents will require many of the same management principles traditionally applied to employees.

As organisations deploy increasingly capable AI systems, leaders will need visibility into how those agents operate, what tasks they perform and how decisions are made. Nadella stressed the importance of transparency, saying businesses must be able to inspect and audit AI systems in detail.

“I first need to know, what are these agents? What are they doing? What are their reasoning traces? They need to be fully inspectable, fully auditable,” Nadella said.

He explained that enterprises will have to assign identities to AI agents, place them within controlled environments and establish governance policies that define what they can access and how they can operate.

The comments reflect a growing debate across the technology industry about AI accountability. While companies are rapidly adopting AI-powered assistants to automate routine work, concerns remain about security, compliance and decision-making transparency.

Nadella revealed that even he regularly works with large numbers of AI coding assistants. However, he acknowledged that managing many AI systems simultaneously can create a significant mental burden, describing the experience as generating a high cognitive load.

Microsoft is already building infrastructure aimed at helping organisations handle this emerging reality. Nadella pointed to tools within the company's ecosystem that provide identity management, security monitoring and data governance capabilities for AI-powered workflows.

According to him, platforms such as Entra, Defender and Purview are intended to help businesses monitor AI activity, secure access permissions and manage information generated by digital agents.

AI could reshape the structure of companies

Beyond governance, Nadella argued that AI represents a deeper transformation of how businesses operate.

“AI is not like a sort of a technology. It's the future of the firm,” he said, suggesting that the technology could fundamentally alter organisational structures and competitive dynamics across industries.

In his view, the real impact of AI will emerge when systems acquire enough industry-specific knowledge to perform increasingly complex tasks. At that point, companies may need to rethink workflows, management practices and even the way value is created.

However, Nadella also issued a warning about the risks associated with widespread AI adoption.

As businesses increasingly rely on AI systems, protecting proprietary information will become critical. He cautioned that organisations must pay close attention to how data flows between employees, digital systems and AI models.

“Because if you leak it, it's a one-way door. You're done in some sense,” Nadella warned while discussing the dangers of exposing valuable intellectual property.

He advocated for an approach where organisations maintain control over their data and computing environments, allowing AI models to operate within secure boundaries rather than exposing sensitive information externally.

The broader message from Nadella was that the AI era will require companies to think beyond productivity gains. As digital workers become a larger part of the workforce, businesses will need new systems for governance, security and accountability. The challenge, he suggested, will not simply be deploying AI agents but ensuring they operate safely, transparently and in ways that protect the knowledge and assets that give organisations their competitive edge.