Claude from Anthropic refers to Microsoft’s Copilot: OpenAI is not the only option

Microsoft was surprised by the announcement of an alliance with Anthropic, which changes the way of its assistant: the company integrates the **Claude** model into Copilot and launches a version capable of running tasks on a username. This movement opens up new possibilities for businesses and users and represents an important diversification in the face of Microsoft’s historic relationship with **OpenAI**.

This was announced by Microsoft

The company introduced what it calls the next generation of Copilot, where users will tend to switch to more AI models to choose according to their needs. In the first phase, **Claude** is added to the main Copilot chat, although he is currently only available to **Frontier** Early Access members.

Microsoft also introduced **Copilot Cowork**, a version of Anthropic aimed at performing administrative and factory tasks within a suite of products.

How Copilot Cowork works

Copilot Cowork is not a simple conversational assistant: it acts as an agent that translates instructions into a work plan and executes it using the user’s files and applications. It is designed to work within **Microsoft 365**, with direct integration into services such as Outlook, Teams and Excel.

Microsoft ensures that the user remains in control: the agent offers validation points to accept changes, can pause execution, and asks for clarification before making relevant decisions.

• Checking and cleaning the calendar: identifies reunions with low courage, conflicts and suggests reprogramaciones.
• Automation of runs and tasks: you can search, summarize and act on reports and archives.
• Preparation for meetings: generates information packs and materials for the team.
• Investigation and Analysis: performs tasks at businesses, collects data and performs basic financial analysis.
• Landing aircraft: Create timelines and practical steps to describe the product.

Implications for businesses and users

For organizations, this news represents a potential productivity improvement when delegating repetitive routines, but also complex management and security decisions: IT teams will tend to revise permissions and governance policies to allow agent access to files, calendars and documents.

From an end-user perspective, Copilot Cowork promises to save time in administrative areas, although the real value will depend on the accuracy of its planes and the control that surveillance systems offer.

Strategic context: why does it matter now?

The connection with Anthropic leads to a key moment: Microsoft was one of the first inverts in **OpenAI** and over the years integrated GPT models into Copilot. Opening the platform to alternatives like **Claude** suggests a strategy to diversify AI providers and reduce dependence on a single partner.

This changes market dynamics – more significant among large models – and can accelerate the emergence of specialized functions in the enterprise environment.

Follow up next week

Points to watch include the availability of Claude and Copilot Cowork within the **Frontier** program, details on privacy and control for administrators, and the valuable models Microsoft is implementing for these advanced features.

In this short space, companies must prepare access policies, evaluate risks and define priority use cases. It is advisable to check with individual users for information on when you can try these functions and conditions.

The integration of Claude and the launch of Copilot Cowork consolidate the evolution of Copilot into more autonomous and operational assistants, planning as many product opportunities as new management rules in the corporate environment.

Similar articles