According to a report by Business Insider, Microsoft is developing an AI model capable of breaching offline or “air-gapped” systems for intelligence agencies. The model, codenamed “Sowa,” is designed to extract sensitive data from isolated computer networks without an internet connection. It utilizes techniques like electromagnetic and acoustic side-channel attacks to gather information from the targeted systems. The report suggests that Sowa could be used by intelligence agencies to infiltrate secure networks and obtain classified data. Microsoft’s involvement in this project raises concerns about the potential misuse of such powerful AI technology for surveillance and espionage purposes. While the company claims to prioritize ethical AI development, the implications of Sowa highlight the need for robust safeguards and oversight to prevent its exploitation by malicious actors.
Microsoft's AI Model Aims to Breach Offline Systems for Intelligence Agencies
Previous post
New AI Data Center Coming to Wisconsin