AI allows hackers to identify anonymous social media accounts, study finds
New research suggests tech behind AI platforms such as ChatGPT makes it easier to perform sophisticated privacy attacks
TL;DR
- Large Language Models (LLMs) make it easier for hackers to identify anonymous social media accounts.
- In tests, LLMs successfully matched anonymous users to their real identities on other platforms using posted information.
- LLMs make sophisticated privacy attacks cost-effective, requiring less expertise.
- Potential uses include government surveillance of dissidents and hackers launching personalized scams or spear-phishing attacks.
- Concerns exist about commercial de-anonymization products and the potential for false accusations.
- LLMs can potentially use data beyond social media, such as hospital records, necessitating higher anonymization standards.
- AI de-anonymization is not always successful, depending on the amount and consistency of shared information.
- Recommendations include platforms restricting data access (rate limits, detecting scraping) and individuals taking more precautions about shared information.