Microsoft rolls out Copilot Cowork more broadly and lets AI models check each other's work

Microsoft expanded Copilot Cowork availability through its Frontier program and added calendar planning and daily briefings.

Microsoft's Researcher tool gained a Critique function that has one AI draft and a second AI review the draft using Anthropic and OpenAI models.

Microsoft claims the new research agent outperforms Perplexity with Claude Opus 4.6 by seven points but did not compare it to GPT-5 Deep Research.

A new Model Council feature lets users compare answers from different models side by side.

All updates are included in Wave 3 of Microsoft 365 Copilot.