Research
153d agoLLMs Expose Online Anonymity at Scale
Recent studies show that large language models can reconstruct user identities from public posts, undermining anonymity across social media, forums, and messaging platforms worldwide. Researchers demonstrate that even minimal text snippets enable inference of personal details, raising concerns for privacy, free expression, and data protection laws.
The Signal
The findings prompt urgent policy reviews by regulators, including the EU and OpenAI.