← feed
NOTABLE
arXiv: information-flow control for agents is bounded by label creep, per limits study
arXiv 2611.10395 studies information-flow control as an agent-policy enforcement mechanism and its label creep failure mode, where monotonically increasing security labels progressively remove utility; it examines the two mitigation routes (system design and label declassification) and what they cannot fix.
released 2026-11-07 · ingested 2026-11-10T12:35Z
Sources
Primary source
Cite this
APA
AI Security Index. (2026). arXiv: information-flow control for agents is bounded by label creep, per limits study. AI Security Index. https://arxiv.org/abs/2611.10395
BibTeX
@misc{aisec-2026-11-10-arxiv-2611-10395-information-flow-control-label-creep-agentic-systems,
title = {arXiv: information-flow control for agents is bounded by label creep, per limits study},
year = {2026},
howpublished = {\url{https://arxiv.org/abs/2611.10395}},
note = {AI Security Index item 2026-11-10-arxiv-2611-10395-information-flow-control-label-creep-agentic-systems}
}