Researchers from ETH Zurich, Anthropic and the Machine Learning Alignment and Theory Scholars program have built an automated AI system that can link pseudonymous online profiles to real identities. Using large language models to analyze writing style, posting patterns and other clues, the system correctly matched up to 68 percent of accounts with 90 percent precision, far outpacing traditional methods. The experiment cost only a few dollars per profile, highlighting a low‑cost barrier for large‑scale deanonymization. The study warns that online anonymity may be less secure than many assume, especially as AI capabilities continue to improve. Read more →

Researchers demonstrated that large language model (LLM) agents can extract identity clues from free‑text data, search the web autonomously, and match those clues to real‑world individuals. In experiments using interview transcripts, Reddit comments, and a large pool of Reddit users, the AI was able to correctly re‑identify a measurable share of participants while maintaining high precision. The findings highlight a growing capability of AI to breach pseudonymity, raising concerns about privacy in online platforms. Read more →

OpenAI CEO Sam Altman announced that the company will revise its contract with the U.S. Department of Defense to explicitly forbid the use of its artificial‑intelligence system for mass surveillance of Americans. In an internal memo shared on X, Altman detailed new language tying the restriction to the Fourth Amendment and other applicable laws, and said he would prefer jail over complying with an unlawful order. The move follows a broader government debate over AI guardrails, pressure on rival Anthropic to drop safeguards, and a recent surge in Anthropic’s popularity after the policy shift. Read more →