A second-opinion check for Copilot research reports, but it needs a licence, the Frontier programme and an admin who allows Anthropic models.

Kenning's take

What it does

Critique is an option in the Researcher agent of Microsoft 365 Copilot. A GPT model writes the research report and a Claude model then reviews it for source quality, completeness and citations before it is delivered. Microsoft announced it in March 2026 alongside Council, which shows several models' answers side by side.

• GPT drafts, Claude reviews and improves the report
• Checks source reliability and citation grounding
• Rubric-based review for completeness
• Default in Researcher when Auto is chosen
• Council view compares several models side by side

Who it suits. Microsoft 365 Copilot users who write research reports from web and work content.

Key facts

Category

Everyday AI assistants

Part of

Microsoft Copilot

Made by

Microsoft

Released

30 March 2026

How to get it

Researcher agent in the Microsoft 365 Copilot app (desktop and web); needs a Microsoft 365 Copilot licence, the Frontier early-access programme, and an admin to allow Anthropic models

Works on

Microsoft 365 Copilot app (desktop and web)

Pricing

Plan	Price	Billed
Microsoft 365 Copilot (business)	$21	user per month, paid yearly (shown at $18 on offer)

source

FAQ

Facts from the maker's own pages and the sources listed. Checked 2026-10-11.