Microsoft Splits Copilot Into Three Apps and Introduces Autopilot Agent With Usage-Based Billing
Microsoft is restructuring its Copilot app into Home, Code, and a new cloud-based agent called Autopilot, while shifting Autopilot and Code to usage-based billing instead of flat-rate pricing.
Microsoft is reorganizing its Copilot application into three distinct sections: Home, Code, and a new autonomous agent named Autopilot. The move marks another significant overhaul of the company's flagship AI assistant as it seeks to differentiate its offerings for general users, developers, and enterprises.
Autopilot is built on OpenClaw and operates continuously in the cloud, according to Microsoft. Unlike a conventional chatbot that responds only when prompted, the agent can monitor Teams channels and complete tasks on its own. This represents a shift toward persistent, proactive AI that works in the background rather than waiting for direct user input.
Alongside the structural changes, Microsoft is also changing how it charges for Autopilot and Code. The company is moving from flat-rate pricing to usage-based billing for these two sections. That means customers will pay according to how much they use the services, rather than a fixed subscription fee. The change signals a further step away from Microsoft's previous AI subsidy model, under which the company absorbed much of the cost to encourage adoption.
The decision to split Copilot into Home, Code, and Autopilot reflects a broader segmentation strategy. Home likely targets everyday users who want assistance with writing, research, and routine tasks. Code is aimed at software developers who need help generating, debugging, and maintaining code. Autopilot, meanwhile, is designed for organizations that want an always-on agent capable of handling repetitive or monitoring-intensive workflows without constant human oversight.
Usage-based billing could have significant implications for how businesses budget for AI tools. Flat-rate pricing offers predictability, but it can also lead to overprovisioning or underutilization. Usage-based models align costs more closely with actual consumption, which may appeal to customers with variable needs. However, it also introduces uncertainty, as costs can fluctuate depending on how heavily the agent is used.
Microsoft's shift away from subsidizing AI usage suggests the company is moving toward a more sustainable commercial model. As AI services scale, the computational costs of running them continuously in the cloud are substantial. By charging based on usage, Microsoft can better recover those costs while giving customers flexibility.
The introduction of Autopilot also raises questions about the role of autonomous agents in the workplace. An agent that monitors Teams channels and completes tasks on its own could reduce administrative burden, but it also requires trust in its decision-making. Organizations will need to consider governance, oversight, and data privacy before deploying such tools broadly.
For now, Microsoft has not detailed exactly which tasks Autopilot can perform or how customers will be able to configure its access to Teams and other services. More information is expected as the rollout proceeds. The restructuring of Copilot into three sections and the adoption of usage-based billing for Autopilot and Code represent a notable evolution in Microsoft's AI strategy, balancing innovation in autonomous agents with a more disciplined approach to pricing.
5
