LLM Attack: Anonymized Users Identified via Text
OwainEvans_UK · x · 2026-08-25
This arXiv paper demonstrates that LLMs can perform at-scale deanonymization attacks.
Core Findings:
- An agent with full internet access can re-identify Hacker News users and Anthropic Interviewer participants with high precision using only pseudonymous profiles and conversations.
- The method matches what would take hours for a dedicated human investigator.
Attack Pipeline:
- Extract Features: Extract identity-relevant features from unstructured text.
- Search Candidates: Find potential matches via semantic embeddings.
- Reason & Verify: Reason over top candidates to verify matches and reduce false positives.
Datasets:
- Hacker News <-> LinkedIn: Uses cross-platform references.
- Reddit Movie Communities: Matches users across different subreddits.
- Time Split: Splits a single user's Reddit history to create two profiles for matching.
Results: LLM-based methods substantially outperform classical baselines, achieving up to 68% recall at 90% precision.
Related event: Researchers Say AI Hallucination Largely Solved by 2026(4 posts)→
More from Safety
- Anthropic red team's first-ever model "abliteration" cut GLM-5.3 refusals from 90% to 3% for ~$4,400 — cephaloform · 2026-10-16
- Anthropic grants METR employee-like access as independent safety evaluators go public — AnthropicAI · 2026-10-16
- APEX-Harm benchmark: script-based injection attacks easily compromise agents in shared workspaces — KAUST · 2026-10-16
- Residual-stream forecasts predict agent violations before they happen, USC study shows — USC · 2026-10-16
- AI safety drama: risk research called a 'company effort,' team lead responds — logangraham · 2026-10-16
- Brave's Leo adds end-to-end encrypted chat with GLM via attested TEE — gnukeith · 2026-10-16