The Whisper Leak attack involves determining the topic of conversation with AI chatbots through TLS traffic analysis.
Security researchers from Microsoft have developed a side-channel attack technique called Whisper Leak, which allows for the classification of request topics to services based on large language models through passive analysis of encrypted traffic transmitted over a TLS connection. Information regarding the size of network packets and the delays between their transmissions was sufficient to determine the topics of requests to AI chatbots with over 98% accuracy. […]
