Agentic AI re-identifies 72% of individuals from simulated mobility traces by cross-referencing public web sources without human intervention.
arXiv preprint arXiv:2503.15552 , year=
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3representative citing papers
RBI-Eval shows LLMs integrate sensitive memory under benign prompts at rates 8.9-82.9% higher than no-memory baselines, with retrieval systems reducing but not eliminating the effect.
Counter Scam is a multiagent LLM system that integrates safe data handling, nine role-specific NLP tasks, and a 185k-case scam corpus, with fine-tuned small models beating commercial LLMs by over 10% on those tasks.
citing papers explorer
-
Agentic AI-Powered Re-Identification: An Emerging, Scalable Threat to Mobility Microdata Privacy
Agentic AI re-identifies 72% of individuals from simulated mobility traces by cross-referencing public web sources without human intervention.
-
When Should Memory Stay Silent: Measuring Memory-Use Boundaries in Memory-Augmented Conversational Agents
RBI-Eval shows LLMs integrate sensitive memory under benign prompts at rates 8.9-82.9% higher than no-memory baselines, with retrieval systems reducing but not eliminating the effect.
-
An LLM-based Chain-of-Response Counter-Scam System
Counter Scam is a multiagent LLM system that integrates safe data handling, nine role-specific NLP tasks, and a 185k-case scam corpus, with fine-tuned small models beating commercial LLMs by over 10% on those tasks.