Microsoft overhauls Copilot with chat focus, coding and persistent agents

By Splitfeed EditorialLast updated Sep 28, 2026, 06:00 UTC

In late September 2026, Microsoft announced a substantial redesign of its Copilot product, rebuilding the interface around chat and coding and introducing a persistent autonomous agent, referred to internally as boom, that continues operating without constant user input. Alongside the interface changes, the company rolled out a usage-based billing model for Copilot, shifting at least part of its AI pricing away from flat subscription fees. Microsoft also quietly dropped the Copilot+ branding from its new laptop lineup, decoupling the software identity from its hardware marketing.

The moves matter for several reasons beyond a typical product refresh. Usage-based pricing signals that Microsoft is betting enterprise customers will consume enough AI capacity to make consumption billing more profitable than subscriptions, a model that could reshape how businesses budget for AI tools. The persistent agent feature pushes Copilot closer to fully autonomous operation, where the software takes actions over time rather than responding to single prompts, raising practical questions about oversight and error correction. Dropping the Copilot+ laptop brand, meanwhile, suggests the earlier hardware-tied marketing strategy did not land as intended, and Microsoft is recalibrating how it positions AI features across its product lines.

No published reactions from the Pro-AI, Anti-AI, or Middle Ground camps had appeared by the time this story was filed. Pro-AI voices would typically welcome the autonomous agent capability and usage-based model as signs that AI is maturing into a genuine productivity platform. Anti-AI observers would be expected to raise concerns about unsupervised agent actions, potential cost unpredictability under consumption billing, and the pace at which automation is being normalized in workplace software. Middle-ground commentators would likely focus on whether the pricing model genuinely delivers value for smaller organizations or primarily benefits large enterprises with scale to absorb variable costs.

The clearest near-term signal to watch is how enterprise customers respond to the usage-based billing structure once it is in wider deployment, particularly whether adoption accelerates or whether cost unpredictability drives pushback. Early data on the persistent agent's error rates and the extent of human oversight built into its design would also go a long way toward shaping how regulators and corporate buyers evaluate the product.