AI can unmask anonymous users across platforms: Study
A new study finds that large language models (LLMs), the same tech behind popular chatbots, can figure out who's behind anonymous posts by analyzing those posts and searching cross-platform public data to link identity-relevant clues.
Researchers at MATS Research and ETH Zurich managed to match Hacker News users to their LinkedIn profiles, spot Redditors in movie forums, and even re-identify Anthropic Interviewer participants, achieving up to about 68% recall at a high precision level (e.g., 90% precision).
Implications for online anonymity
This means staying anonymous online is getting harder.
The researchers warn that LLMs make it much easier for bad actors to uncover identities, which could lead to more targeted scams or unwanted surveillance.
They say it's time to rethink how we protect privacy on the internet — especially if you rely on pseudonyms or think your posts are private.