Richard Lawler is a senior journalist who follows news in tech, culture, politics, and entertainment. He joined The Verge 2021, after several years of covering news for Engadget.
After Google’s and OpenAI’s AI news on Tuesday, Microsoft followed up with its own announcements. These include details of two “deep-reasoning” agents for Microsoft 365 Copilot, which it claims are the world’s first. They are called Researcher and Analyst, and they also have new capabilities for custom AI agent. Researcher uses OpenAI’s deep-research AI model, along with connectors to third-party sources like Salesforce and ServiceNow to provide insights across business tools.
The o3 mini reasoning model is used by Analyst, and Microsoft claims it can perform calculations, run Python code, and produce reports at the level of an expert data scientist.
These tools will be available in April for Microsoft 365 Copilot users in an early-access program. They will also include new autonomous agent capabilities, which are now being rolled out in Copilot Studio. Microsoft claims that the new agent flows for Copilot can “automate any task” with rule-based workflows and AI actions. The LinkedIn announcementdescribes scenarios such as an agent flow that sends feedback emails to a specific team. We’ll have to test it out to see if it’s any better than adding checkboxes or two.
