Researchers from ETH Zurich, MATS, and Anthropic built an AI pipeline that matched anonymized Hacker News accounts to real LinkedIn profiles at 90% precision. This attack requires no hacking or data breach, relying solely on web search, embeddings, and reasoning models at a cost of $1 to $4 per target. In a test on 338 Hacker News users who had linked their LinkedIn profile, the AI correctly identified 67% of individuals, while a separate test on 125 scientists allowed it to recognize at least 9 people based solely on how they described their work. The researchers stated that practical online anonymity no longer exists and that threat models for online privacy must be reconsidered. The code, prompts, and real identities uncovered were kept confidential, and the study was approved by ETH Zurich’s ethics review board.
Source: Read the original article

