Pith. sign in

REVIEW 5 cited by

Large Language Models are Advanced Anonymizers

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2402.13846 v2 pith:BGLYOU5Z submitted 2024-02-21 cs.AI cs.CLcs.CR

classification cs.AIcs.CLcs.CR
keywords anonymizationadversarialanonymizersllmstextscapabilitiesevaluationhuman
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Recent privacy research on large language models (LLMs) has shown that they achieve near-human-level performance at inferring personal data from online texts. With ever-increasing model capabilities, existing text anonymization methods are currently lacking behind regulatory requirements and adversarial threats. In this work, we take two steps to bridge this gap: First, we present a new setting for evaluating anonymization in the face of adversarial LLM inferences, allowing for a natural measurement of anonymization performance while remedying some of the shortcomings of previous metrics. Then, within this setting, we develop a novel LLM-based adversarial anonymization framework leveraging the strong inferential capabilities of LLMs to inform our anonymization procedure. We conduct a comprehensive experimental evaluation of adversarial anonymization across 13 LLMs on real-world and synthetic online texts, comparing it against multiple baselines and industry-grade anonymizers. Our evaluation shows that adversarial anonymization outperforms current commercial anonymizers both in terms of the resulting utility and privacy. We support our findings with a human study (n=50) highlighting a strong and consistent human preference for LLM-anonymized texts.

Discussion (0). Sign in to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. PromptPET: Privacy-Utility Optimized Prompt Obfuscation

    cs.CR 2026-07 conditional novelty 7.0 of 10

    PromptPET selectively applies four obfuscation actions (including novel noising) via an OPRO-style rule optimizer to match single-action privacy-utility frontiers and outperform prior prompt-minimization methods on Wi...

  2. Retrieval-Confused Generation is a Good Defender for Privacy Violation Attack of Large Language Models

    cs.CR 2025-06 conditional novelty 6.0 of 10

    A retrieval-confused generation defense that swaps query comments with the most irrelevant paraphrased comments, reducing privacy attack success rates across eight LLMs.

  3. Unveiling Unicode's Unseen Underpinnings in Undermining Authorship Attribution

    cs.CR 2025-08 unverdicted novelty 4.0 of 10

    The paper proposes integrating Unicode steganography into adversarial stylometry to undermine authorship attribution.

  4. PBa-LLM: Privacy- and Bias-aware NLP using Named-Entity Recognition (NER)

    cs.CL 2025-06 conditional novelty 4.0 of 10

    NER-based removal of person and location entities from resumes preserves occupancy-prediction accuracy on FairCVdb, and combined with a debiasing module yields gender-balanced shortlists.

  5. SoK: The Privacy Paradox of Large Language Models: Advancements, Privacy Risks, and Mitigation

    cs.CR 2025-06 conditional novelty 3.0 of 10

    A systematization-of-knowledge survey that categorizes LLM privacy risks into training data, prompts, outputs, and agents, and reviews limitations of current mitigations.

Pith tools