Microsoft Copilot Introduces Multi-Model Intelligence

Microsoft has introduced two new multi-model AI capabilities, Critique and Council, to its Microsoft 365 Copilot Researcher agent, announced on March 30, 2026.

What Is the Critique Feature?

The headline feature, "Critique," equips Copilot's Researcher agent to draw on outputs from both OpenAI's GPT and Anthropic's Claude models for each response.

• GPT is responsible for generating the initial answer
• Claude evaluates that response for accuracy and overall quality

Then the final output gets delivered to the user.

Microsoft said it plans to further evolve this system into a bi-directional workflow, where both models can review and refine each other's outputs. The Critique feature is built around a rubric-based evaluation. It does not take up the rewriting task, rather It provides feedback for the researcher to improve the work.

It will be the default experience in Researcher when "Auto" is selected in the model picker. The review focuses on three key dimensions:

• Source reliability: Verifying where information comes from
• Report completeness: Ensuring nothing critical is missing
• Citation accuracy: Making sure all key claims are backed by verifiable sources

The results show that researchers now score 13.8% higher on the DRACO (Deep Research Accuracy, Completeness, and Objectivity) benchmark, the industry standard for deep research quality.

What Is the Council Feature?

Council takes a different approach.

• Instead of one model reviewing another, it runs an Anthropic and OpenAI model simultaneously, with each producing a complete, standalone report.
• Once both reports are generated, a dedicated judge model evaluates them and creates a distilled summary highlighting where the models meaningfully agree or diverge.

CEO Satya Nadella announced the features on X, writing: "New in M365 Copilot Council. You can run multiple models on the same prompt at the same time, so you can see where they align and diverge."

If you want to run multiple AI models side by side, Chatly also lets you access GPT, Claude, and other leading models in one workspace.

Copilot Cowork Now in Early Access

Alongside the Researcher updates, Microsoft announced that Copilot Cowork is now available through its Frontier early access program.

Cowork handles long-running, multi-step tasks autonomously, covering functions like:

• Calendar management
• Budget reviews
• Meeting preparation

Nicole Herskowitz, corporate vice president of Microsoft 365 and Copilot, said: "Having various different models from different vendors in Copilot is highly attractive, but we're taking this to the next level, where customers actually get the benefits of the models working together."

Adoption Still a Challenge

As of January 2026, Microsoft reported 15 million paid Copilot seats, roughly 3.3% of its 450 million commercial Microsoft 365 users. Features like Critique and Council appear designed to accelerate adoption by addressing the most common objection to AI-assisted research: trustworthiness.

Suggested Reads

• OpenAI Closes Record $122 Billion Funding Round at $852 Billion Valuation
• Anthropic Launched Claude Opus 4.5 — New Flagship Model for Coding and Complex AI Workflows
• OpenAI Prepares to Make a Historic Public Market Debut, Eyeing Up to $1 Trillion Valuation
• ChatGPT Voice Mode Update for Seamless Integration of Voice, Text & Live Maps