Why does TLS not actually hide what I am asking my AI assistant?
机会
Streaming LLM responses emit tokens one at a time over an encrypted connection, and each packet arrival changes size and timing in ways that correlate with topic. The Whisper Leak paper published in November 2025 demonstrated topic classification with above 98% AUPRC across 28 major providers using only encrypted traffic metadata. Several providers issued patches but each mitigation trades query privacy for added latency and there is no agreed specification for what a sufficient defense looks like. The structural problem is that any low-latency streaming protocol leaks information through timing, so padding and noise are palliatives rather than solutions.
为什么重要
Every enterprise using a hosted LLM for sensitive work is operating under a false assumption about what TLS protects, and the fix requires rethinking the streaming transport layer itself.
我如何评估机会
机会评分是我的个人判断，而非量化指标：痛苦程度、发生频率，以及当前解决方案的匮乏程度。分数越高，意味着我认为越值得去构建。
出现时造成的痛苦程度。
人们实际遇到它的频率。
当前针对它的优质工具有多匮乏。