AI Unmasks Anonymous Users: New Study Flags Major Online Privacy Threat

• A new study reveals that large language models (LLMs) like ChatGPT can identify real-world identities of anonymous social media users by analyzing small fragments of personal information.
• Researchers Simon Lermen and Daniel Paleka demonstrated how AI can link anonymous accounts to known online identities using publicly available data, such as anecdotes or location hints (e.g., "Dolores Park" and dog "Biscuit").
• This capability significantly reduces the cost and complexity of privacy attacks, enabling large-scale identity tracing and forcing a "fundamental reassessment of what can be considered private online."
• Experts warn of risks including governments tracking activists, cybercriminals creating personalized scams, and the ease with which even inexperienced attackers can now conduct advanced data analysis.
• While AI linking isn't flawless and can make mistakes, the findings urge individuals and organizations to rethink online anonymity, limit data scraping, and be more cautious about shared personal details.