Microsoft has added multimodel features to its Copilot research assistant to let users run more than one AI system in a single workflow as it tries to lift adoption of Copilot.
The company introduced a feature called Critique that pulls outputs from OpenAI’s GPT and Anthropic’s Claude models for each response, with GPT drafting and Claude reviewing before results are shown.
Microsoft also launched Model Council, which lets users compare outputs from different models side by side, and expanded access to its Copilot Cowork tool for early-access customers.
A Microsoft executive said using multiple vendors’ models together can help keep in check AI hallucinations and improve output quality amid competition from Google Gemini and other agent products.
🔗 Source: Reuters
Originally written by: Diya Lal
Source: Tech in Asia
Published on: 30 March 2026
Link to original article: Microsoft rolls out Copilot Cowork, AI upgrades
