A local multi-agent firewall intercepts LLM web and API traffic and detects sensitive data with hybrid detectors, reaching up to 94.93% F1 on a PII benchmark.
The nightfall ai platform
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Multi-Agent Firewall Architecture for Privacy Protection of Sensitive Data in Interactions with Language Models
A local multi-agent firewall intercepts LLM web and API traffic and detects sensitive data with hybrid detectors, reaching up to 94.93% F1 on a PII benchmark.