Key Points:
- Microsoft is introducing new Copilot and agentic experiences on Windows 11 to make AI-powered agents more productive and secure.
- The new feature, Copilot Actions, allows AI agents to take real actions on behalf of users, such as booking a table or ordering groceries, while keeping them informed and in control.
- Microsoft is committed to including robust security and privacy controls to empower customers to explore the potential of agentic AI confidently.
Microsoft announced new Copilot and agentic experiences on Windows 11, making powerful AI easy to use. Copilot Actions, an experimental feature, allows AI agents to take actions directly on local files in Windows, building on the announcement in May of Copilot Actions on the web. This feature will be previewed soon for Windows Insiders in Copilot Labs.
Copilot Actions is an AI agent that completes tasks for users by interacting with their apps and files, using vision and advanced reasoning to click, type, and scroll like a human would. This transforms agents from passive assistants into active digital collaborators that can carry out complex tasks to enhance efficiency and productivity. When integrated with Windows, the agent can take advantage of what users already have on their PC, like apps and data, to complete tasks.
However, agentic AI introduces novel security risks, such as cross-prompt injection (XPIA), where malicious content can override agent instructions, leading to unintended actions like data exfiltration or malware installation. To address these challenges, Microsoft is establishing a set of security and privacy principles, including distinct agent accounts, limited agentic privileges, operational trust, and privacy-preserving design.
The company is creating the ability for agents in Windows to operate with dedicated agent accounts, distinct from the user account on the device. Agents will start with limited permissions and will only obtain access to resources users explicitly provide permission to. Operational trust will ensure that agents that integrate with Windows are signed by a trusted source, allowing maliciously or poorly behaved agents to be revoked and blocked.
Microsoft is also introducing four new building blocks to Windows 11 to support the exploration of agentic AI, including user control, agent accounts, agent workspace, and user transparency. These building blocks will provide a contained environment for agents to work in parallel with human users, enabling runtime isolation and granular permissions.
As Microsoft expands agentic capabilities in Windows, the company will continue to evolve its defenses. With the upcoming preview release of Copilot Actions to Windows Insiders in Copilot Labs, Microsoft looks forward to gathering valuable feedback that will help shape the experience further ahead of broader release. The Windows platform and its security controls will also be available for other developers in private preview soon to test and provide input.
Windows 11 is the most secure version of Windows ever built, and as the company enters this new agentic era, its commitment is clear: Windows will be the most secure, trusted, and user-centric platform for agentic computing. Microsoft will share more information at Microsoft Ignite 2025 in November.
Read the rest: Source Link
You might also like: Try AutoCAD 2026 for Windows, best free FTP Clients on Windows & browse the best Surface Laptops to buy.
Remember to like our facebook and our twitter @WindowsMode for a chance to win a free Surface every month.