Microsoft has issued a warning to Windows 11 users about a new AI-powered feature that is still undergoing testing on computer systems.
The company explained that the “Proxy Server” feature, which relies on artificial intelligence, appeared in build 26220.7262 of Windows 11. It can be activated manually from the “AI Components” section. Upon activation, the system immediately displays a warning indicating that the technology is still experimental and may affect device performance—ranging from incorrect responses to potential malfunctions in system behavior.
اضافة اعلان
Microsoft clarified that the foundational language model powering the new “Proxy Server” remains under development, which means it may produce answers that appear convincing but are inaccurate. This is due to incomplete training data, misinterpretation of context, and other contributing factors.
Technology experts note that some cybercriminals have already begun testing ways to target such new AI-related features. One of the primary threat scenarios involves the covert replacement of instructions within the “server,” where malicious commands can be embedded in documents or UI elements and executed automatically by the AI. In such cases, the server could theoretically install malware or leak sensitive data.
Further complicating the situation is the fact that the new server is, by default, granted read and write permissions to key user folders such as Documents, Desktop, Downloads, and others.
Microsoft promises to strengthen protection by adding more precise permission controls and safeguards against potential cyberattacks, and currently recommends enabling this experimental feature only for users who fully understand the risks involved.
