The potential privacy risk of knowing an AI approach worked
The author of a privacy primer says discussion of Navier–Stokes results prompted a look at OpenAI, Anthropic, Google DeepMind and xAI policies on training, retention, feedback and chat deletion.
TLDR
The primer’s author argues that simply knowing someone made an approach work with AI can change where others put their time and compute—even without seeing the proof, the person’s identity or their chats. In the author’s view, hiding raw conversations through aggregation or privacy filters does not settle what others can learn.
Combined views
12.1K
6 Sources, first seen 8h ago
