Microsoft repositionne Copilot face à OpenAI et Anthropic : la phase de l'ami-ennemi se termine

Microsoft is now openly pitching Copilot against OpenAI and Anthropic, per Tech in Asia (16 July 2026). After the Build 2026 orchestration layer, Redmond ends the wrapper era and prices its own model story as the differentiator.

In plain terms. Microsoft is pitching Copilot as a peer to OpenAI and Anthropic assistants, not as a distribution channel for them. In the same year it shipped the Build 2026 agent orchestration layer, Redmond ends the frenemy phase.

Our analysis

The trajectory was visible: Microsoft's spend on OpenAI training runs did not buy exclusivity, and MAI-family and Phi models have been quietly maturing. Positioning Copilot as a competitor is the pricing wedge - enterprises paying for Microsoft 365 already have Copilot licensing and don't need a second contract with OpenAI. This sits inside the saas-is-over thread: when the intelligence layer replaces feature depth, the vendor that owns the seat (Microsoft) captures the AI monetisation the platform vendor (OpenAI) can't.

To watch

(1) Copilot model composition - does OpenAI-model share fall in the next Copilot Enterprise SKU? (2) Data-residency contract language - Copilot's "your data stays yours" now becomes the anti-OpenAI axis. (3) Anthropic's IPO S-1 (October, art #21622915) will show how much revenue depends on Microsoft-adjacent enterprise deals. Three data points, same tectonic shift.