Microsoft is introducing a “deep research” AI-powered tool in Microsoft 365 Copilot, its AI chatbot app.
There’s been a raft of deep research agents launched recently across chatbots, including OpenAI’s ChatGPT, Google’s Gemini, and xAI’s Grok. Powering them are so-called reasoning AI models, which posses the ability to think through problems and fact-check themselves — skills arguably important for conducting in-depth research on a subject.
Microsoft’s flavors are called Researcher and Analyst.
Researcher combines OpenAI’s deep research model — which powers the company’s own ChatGPT deep research tool — with “advanced orchestration” and “deep search capabilities.” Microsoft claims that Researcher can perform analyses including developing a go-to-market strategy and creating a quarterly report for a client.
As for Analyst, it’s built on OpenAI’s o3-mini reasoning model and is “optimized to do advanced data analysis,” Microsoft said. Analyst progresses through problems iteratively, taking steps to refine its “thinking” and provide a detailed answer to queries. Analyst can also run the programming language Python to tackle complex data queries, Microsoft added, and expose its “work” for inspection.
What makes Microsoft’s deep research
tools slightly more unique than the competition is their access to work data as well as the worldwide web. For example, Researcher can tap third-party data connectors to draw on data from AI “agents,” tools, and apps like Confluence, ServiceNow, and Salesforce.
Granted, the real challenge is ensuring tools such as Researcher and Analyst don’t hallucinate or otherwise make stuff up. Models including o3-mini and deep research are by no means perfect; from time to time, they mis-cite work, draw incorrect conclusions, and pull from dubious public websites to inform their reasoning.
Microsoft is launching a new Frontier program through which Microsoft 365 Copilot customers can gain access to Researcher and Analyst. Those enrolled in Frontier, which going forward will gain experimental Copilot features first, will get Researcher and Analyst starting in April.
Kyle Wiggers is TechCrunch’s AI Editor. His writing has appeared in VentureBeat and Digital Trends, as well as a range of gadget blogs including Android Police, Android Authority, Droid-Life, and XDA-Developers. He lives in Manhattan with his partner, a music therapist.
Subscribe for the industry’s biggest tech news