On Monday, Microsoft expanded its enterprise AI efforts by giving more early-access customers the chance to use Copilot Cowork.
The company also introduced new Copilot features that let users combine results from both OpenAI’s GPT and Anthropic’s Claude models in one workflow.
This step shows Microsoft’s ongoing effort to make Copilot more helpful for workplace tasks and to boost its use as competition in enterprise AI grows.
Copilot Cowork moves beyond limited testing
This wider release is a step forward from Microsoft’s earlier testing phase. Reuters reported that Copilot Cowork was first introduced in testing earlier this month.
They described it as a tool based on Anthropic’s Claude Cowork product, created in response to the rising demand for autonomous AI agents.
ITPro noted that, until now, only a small group of customers could use Cowork with Copilot.
Now, the broader rollout is giving more users access through Microsoft’s Frontier program, which offers early AI features.
Microsoft puts GPT and Claude in the same loop
The biggest update is a new feature called Critique. According to Reuters, Copilot’s Researcher agent can now use outputs from both OpenAI’s GPT and Anthropic’s Claude models for each response.
This means it no longer relies on just one model. In this setup, GPT creates the first draft, and Claude checks it for accuracy and quality before the user sees the answer. Microsoft plans to make the process two-way in the future, so GPT can review Claude’s drafts too.
Nicole Herskowitz, corporate vice president of Microsoft 365 and Copilot, told Reuters that having various different models from different vendors in Copilot is already appealing. However, Microsoft wants customers to benefit from these models working together.
Microsoft believes this multi-model approach can make workflows faster, reduce AI errors, and give more reliable results.
This should improve both productivity and quality.
Model Council adds side-by-side comparisons
Microsoft is also launching Model Council. This feature lets users compare responses from different AI models side by side.
This shows that Microsoft wants Copilot to be seen as more than just a single-model assistant. It gives enterprise users a way to review and compare outputs in one system.
What Copilot Cowork is designed to do
ITPro, referencing a Microsoft blog post, reported that Jared Spataro said Copilot Cowork will help teams automate tasks and increase productivity.
The publication quoted Spataro, who said Copilot Cowork makes it easy to delegate and complete work and can create a plan, reason across your tools and files, and carry work forward with visible progress and opportunities to steer.
Claude Cowork, launched in January, is a workflow tool that lets users automate tasks with AI agents, using plugins for marketing, legal, and data analysis.
Copilot Cowork will be available through Microsoft 365’s E7 AI subscription tier once it is fully released.
The publication also quoted Capital Group executive Barton Warner, who said the new Cowork features would help the company automate and scale the Copilot ecosystem.
He added that the tool is about taking real action by connecting steps, coordinating tasks, and following through on daily workflows.
Microsoft’s bigger enterprise AI play
This rollout comes as Microsoft faces more competition in enterprise AI from Google’s Gemini and other new agentic products.
The company has been working quickly to boost Copilot adoption as it competes with rivals and autonomous agents like Claude Cowork.
Monday’s updates show that Microsoft believes enterprise customers want more than just a chatbot that gives one answer.
The company now offers a Copilot experience where several models can draft, review, compare, and coordinate work within a single system.
