Microsoft is rolling out Critique, a multi-model deep research system integrated into M365 Copilot, Satya Nadella said in a LinkedIn post announcing the feature.
The tool is designed to combine multiple models to produce, evaluate and refine research outputs and reports, marking a shift toward model orchestration rather than reliance on a single large model.
Related reading
- Microsoft warns AI agents risk becoming "double agents" as it unveils security controls at RSAC
- Nvidia says the open versus proprietary AI debate is the wrong argument
- Meta hires founders of AI startup Dreamer in latest push to build autonomous agent technology
Satya Nadella wrote, "Introducing Critique, a new multi-model deep research system in M365 Copilot," and added that "You can use multiple models together to generate optimal responses and reports," the post said. Early reactions on the same thread praised the approach for improving depth and confidence in research workflows, while other commenters raised governance concerns about reviewer model oversight.
One commenter highlighted that capability and accountability must scale together, asking explicitly, "who audits the auditors?" and urging clarity on reviewer rubrics and provenance. Microsoft said the feature is "Available today in Frontier," indicating immediate availability for Frontier customers.
The recap
- Microsoft launches Critique, multi-model research system in M365 Copilot
- It enables multiple models to generate and critique responses
- Satya Nadella says Critique is "Available today in Frontier"
