Microsoft 365 Copilot’s Researcher Agent Now Uses GPT and Claude to Improve Answers

Microsoft pairs GPT and Claude to improve Copilot research and workflows.

Hero approved Microsoft 365

Key Takeaways:

  • Microsoft added a new Critique feature to the Microsoft 365 Copilot Researcher agent.
  • The Researcher agent now supports both GPT and Claude models.
  • Microsoft also launched a Claude‑powered Copilot Cowork agent that turns goals into coordinated, multi‑step actions across apps.

Microsoft has announced a major upgrade to Microsoft 365 Copilot’s Researcher agent. This update brings a new Critique feature that leverages OpenAI’s GPT and Anthropic’s Claude models to enhance the quality of its responses.

Microsoft 365 Copilot’s Researcher agent is an AI capability designed to help users answer complex questions by pulling together information from multiple sources and turning it into clear, actionable analysis within their work environment. It goes beyond simple search by synthesizing data, structuring insights, and providing reasoned responses with citations. This agent uses multiple AI models to draft, review, and improve accuracy and completeness, which makes it feel like having a team of analysts supporting deep research and decision‑making directly inside Microsoft 365.

In September, the Researcher agent added support for Claude models as an alternative to GPT models. This new Critique mode will allow the AI agent to use both Claude and GPT in the same task.

“GPT drafts, Claude reviews for accuracy, completeness, and citation integrity before it’s delivered. We expect this workflow to be bidirectional in the future. The results are measurable—Researcher now scores 13.8% higher on the Deep Research Accuracy, Completeness, and Objectivity, or DRACO benchmark, the industry standard for deep research quality,” Microsoft explained.

microsoft researcher agent
Critique Mode (Image Credit: Microsoft)

Model Council enables side‑by‑side AI comparisons

The Researcher agent also includes a model comparison feature (model Council), which allows users to see side‑by‑side responses from different AI models to evaluate agreements, differences, and unique perspectives.

Microsoft noted that the Researcher agent is currently available to customers with a Microsoft 365 Copilot subscription. The company has also announced a new Claude‑powered Copilot Cowork agent, which is designed to handle long‑running, multi‑step work by turning high‑level goals into coordinated action across apps, files, and workflows.

Copilot Cowork incorporates Claude’s reasoning strengths with Microsoft’s enterprise skills to create structured plans, execute steps over time, track progress, and adapt as work evolves. It helps security teams delegate and complete complex tasks more efficiently.