tech
March 3, 2026
LLMs can unmask pseudonymous users at scale with surprising accuracy
Pseudonymity has never been perfect for preserving privacy. Soon it may be pointless.
TL;DR
- AI-powered research can deanonymize pseudonymous social media users with high recall (up to 68%) and precision (up to 90%).
- Large Language Models (LLMs) invalidate the assumption that pseudonymity provides adequate online privacy, as they can identify users from free text and web interactions.
- These techniques outperform classical deanonymization methods, which previously required structured data and extensive manual effort.
- Potential consequences include increased doxxing, stalking, hyper-targeted advertising, and social engineering scams.
- Mitigation strategies include platform rate limits, detection of automated scraping, restricted data exports, and LLM providers building guardrails against misuse.
Continue reading the original article