Microsoft warns Windows 11 AI could put malware on your PC

Microsoft keeps injecting AI into Windows, and now the company itself is admitting that there are security risks in doing so.

This week, Microsoft added some new Agentic AI features for Windows 11 Insider users, allowing AI to automate things like sending emails and sorting files. These are turned off by default and need to be selected, but for those who choose to enable them, Microsoft has published a security note on its website warning that there are security risks in giving AI access to all your files:

“As these capabilities are introduced, AI models still face functional limitations in terms of their behavior and may sometimes hallucinate and produce unexpected output,” Microsoft said. “Additionally, agentic AI applications introduce new security risks, such as cross-prompt injection (XPIA), where malicious content embedded in UI elements or documents can override agent instructions, leading to unintended actions such as data exfiltration or malware installation.”

In other words, it is technically possible for something that is intended to help to cause harm to users. That could very well happen Very good An unlikely hypothetical edge case, but the fact that Microsoft felt compelled to say anything about it is a little worrying. As a potential solution, Microsoft is introducing an experimental feature called “Agent Workspaces”, which limits AI agent access on PCs. In basic terms, this means that the agent can only access things that are available to any user of the machine, while files locked behind specific user profiles are off-limits.

We’re still in the relatively early stages of all this, so it’ll take some time to see how this pans out. But be careful before turning on these features.