Microsoft on Friday introduced Autopilot, a revamped Copilot platform that brings chat, coding and agent-based capabilities together into a single workplace tool. The company says the system will support tasks ranging from assembling presentations and building spreadsheets to analyzing data and surfacing connections users might otherwise miss.
Two billing options
Autopilot will be offered with two billing models. A per-user subscription is intended for general-purpose AI tasks and is subject to a token limit that Microsoft says most subscribers will never hit. The alternative is a usage-based option tailored for longer, multi-step agent workflows.
Satya Nadella, Chief Executive Officer of Microsoft, emphasized at the unveiling that the platform’s success will hinge on user trust: organizations and individuals must feel comfortable delegating credentials and autonomous actions to the system, and they must retain a sense of control over outcomes.
Model routing and providers
Microsoft said Autopilot will automatically route user requests to the most appropriate model from offerings by OpenAI and Anthropic, while also allowing manual model selection. The company added that it plans to make additional providers’ models available on the platform in the future.
Why it matters
By consolidating multiple Copilot-style features into one interface and offering flexible billing, Autopilot aims to simplify enterprise AI adoption. The dual pricing approach may be especially useful as model developers and providers gradually change how they support customers, giving organizations the choice between predictable per-user subscriptions and consumption-based billing for heavier agent-driven usage.
Additional notes
Microsoft highlighted user control and trust as central considerations for adopting Autopilot. An AI assistant contributed to preparing this article; the final content was edited and verified by a journalist.
