Computer-using agents now deliver more secure UI automation at scale
See how new updates to computer‑using agents improve UI automation with secure credentials, detailed monitoring, and scalable Cloud PC capacity.
Get an in-depth first look at the newest Copilot Studio features.
See how new updates to computer‑using agents improve UI automation with secure credentials, detailed monitoring, and scalable Cloud PC capacity.
xAI models are now available in Copilot Studio, expanding your multi‑model lineup with a new option for fast reasoning and flexible agent design.
Today we’re excited to bring OpenAI’s GPT-5.2 to Microsoft 365 Copilot and Microsoft Copilot Studio.
Claude Opus 4.5 is now in Copilot Studio, offering sharper reasoning, better long‑context, and stronger multi‑step performance for your agents.
Today, we are making GPT-5.1 available in Microsoft Copilot Studio, alongside OpenAI’s release. This is available as an experimental model for U.S. customers in early release cycle Power Platform environments. The GPT-5.1 series brings improved adaptability in thinking time in both chat and reasoning.
In this edition of our monthly roundup, we’re recapping new features released in Microsoft Copilot Studio in October 2025.
Automated agent testing is now built into Copilot Studio—evaluate performance, improve quality, and scale confidently with Agent Evaluation.
In this edition of our monthly roundup, we’re recapping new features released in Microsoft Copilot Studio in September 2025.
Claude Sonnet 4.5 – Anthropic’s latest and most capable model – is rolling out today in Microsoft Copilot Studio.
Anthropic models are now available in Copilot Studio, providing more flexibility to design smarter agents, speed up workflows, and improve outcomes.
Computer use is now available in public preview in US-based environments, expanding how you can design agents that work across websites and apps.
For organizations that need deeper oversight and real-time control, a new feature is now in public preview: Advanced real-time protection during agent runtime for enhanced security.