PDF.

We show that large language models can be used to perform at-scale deanonymization. With full Internet access, our agent can re-identify Hacker News users and Anthropic Interviewer participants at high precision, given pseudonymous online profiles and conversations alone, matching what would take hours for a dedicated human investigator. We then design attacks for the closed-world setting. Given two databases of pseudonymous individuals, each containing unstructured text written by or about that individual, we implement a scalable attack pipeline that uses LLMs to: (1) extract identity-relevant features, (2) search for candidate matches via semantic embeddings, and (3) reason over top candidates to verify matches and reduce false positives. Compared to prior deanonymization work (e.g., on the Netflix prize) that required structured data or manual feature engineering, our approach works directly on raw user content across arbitrary platforms. We construct three datasets with known ground-truth data to evaluate our attacks. The first links Hacker News to LinkedIn profiles, using cross-platform references that appear in the profiles. Our second dataset matches users across Reddit movie discussion communities; and the third splits a single user’s Reddit history in time to create two pseudonymous profiles to be matched. In each setting, LLM-based methods substantially outperform classical baselines, achieving up to 68% recall at 90% precision compared to near 0% for the best non-LLM method. Our results show that the practical obscurity protecting pseudonymous users online no longer holds and that threat models for online privacy need to be reconsidered.

  • CerebralHawks@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    9
    ·
    21 hours ago

    It is absolutely possible to identify users who post a lot on a public forum with a real name (e.g. Facebook or the like) as well as Reddit. So say you have some politician who claims to have X, Y, Z values and a Reddit user who has A, B, and C values that are antonymous to X, Y, and Z. By comparing common phrases, as well as by charting when the two seemingly separate users are online, you could say with reasonable certainty that the two people are one and the same, especially if you prompt them carefully to say the kinds of things they would say about neutral topics on both accounts. It would be hard to get 100% certainty, but you’d be close enough to imply it’s them.

    AIs (LLMs) just make it faster.

    Don’t post about controversial politics if you also post under your real name. It’s not a matter of “mask yourself better.” There will always be tells.

    • LwL@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 hour ago

      I’ve always acted assuming this to be possible, but it used to require either an unhinged individual or some other reason for a very dedicated investigation. The barrier being potentially that much lower is scary, particularly for anyone with a bit of internet fame that would rather stay anonymous