Satya Nadella Says AI Agents Could Create a Market Bigger Than Cloud

Microsoft CEO Satya Nadella has predicted that artificial intelligence agents could eventually create a market “orders of magnitude” larger than cloud computing. Nadella made the comments in a recent discussion about Microsoft’s expanding AI strategy, arguing that the technology is moving beyond traditional chatbot interactions toward systems capable of performing tasks independently on behalf of users and organisations.

The comments come as Microsoft continues to develop Copilot around increasingly autonomous AI agents. The company recently introduced Autopilot, a persistent agent designed to continue working on assigned tasks without requiring a new instruction every time. Microsoft says the system can monitor work channels, follow up on conversations, carry out recurring assignments and resume projects over time, subject to organisational permissions and controls.

Nadella’s vision reflects a broader change in how Microsoft expects people to interact with software. Instead of simply asking an AI system a question and receiving an answer, users could delegate entire workflows to agents. Microsoft’s new Copilot platform combines conversational AI with tools for building applications, working with Office documents and automating business processes, allowing AI to become more deeply integrated into everyday workplace operations.

The shift also has major implications for Microsoft’s cloud business because autonomous agents require computing resources to operate continuously. Microsoft has reported strong growth in Azure and increasing adoption of its tools for building AI agents. The company has said customers using its Fabric and Foundry products have grown significantly, while demand for cloud computing capacity has continued to increase as businesses experiment with agent-based AI systems.

Nadella has also acknowledged that the expansion of AI agents creates new risks that companies must manage. He has called for stronger testing, transparency and human control, particularly because autonomous systems can make unexpected decisions while performing complex tasks. His comments suggest that the future of AI will involve not only more capable agents, but also systems for monitoring, securing and governing them as they take on a greater share of workplace activities.