Coding agents struggle to infer least-privilege file permissions by omitting needed accesses while granting unused or sensitive ones, but Sufficiency-Tightness Decomposition improves sensitive-task success by up to 15.8% and reduces attacks.
A vision for access control in llm-based agent systems
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
fields
cs.CR 4years
2026 4representative citing papers
Execution-security research for AI coding agents is fragmented across 17 mechanism categories with five unaddressed cross-cutting gaps, including missing head-to-head isolation-vs-capability evaluation and untested real-world denylist fragility.
Proposes an OS-centered privacy framework for on-device AI that treats privacy as institutional accountability, including a threat model, six-part risk taxonomy, privacy-by-architecture controls, and four-level audit rubric demonstrated on Apple, Android, and Microsoft systems.
A data-centric survey finds that only information-flow control covers compositional and cross-session leakage in LLM agents and that no single benchmark tests an agent across all its data surfaces under one policy.
citing papers explorer
-
Do Coding Agents Understand Least-Privilege Authorization?
Coding agents struggle to infer least-privilege file permissions by omitting needed accesses while granting unused or sensitive ones, but Sufficiency-Tightness Decomposition improves sensitive-task success by up to 15.8% and reduces attacks.
-
The Balkanization of Execution-Security Research for AI Coding Agents: Isolation, Access Control, and Time-of-Check-to-Time-of-Use Vulnerabilities
Execution-security research for AI coding agents is fragmented across 17 mechanism categories with five unaddressed cross-cutting gaps, including missing head-to-head isolation-vs-capability evaluation and untested real-world denylist fragility.
-
Local Is Not a Sufficient Privacy Boundary: Governing OS-Integrated On-Device AI
Proposes an OS-centered privacy framework for on-device AI that treats privacy as institutional accountability, including a threat model, six-part risk taxonomy, privacy-by-architecture controls, and four-level audit rubric demonstrated on Apple, Android, and Microsoft systems.
-
Agents That Know Too Much: A Data-Centric Survey of Privacy in LLM Agents
A data-centric survey finds that only information-flow control covers compositional and cross-session leakage in LLM agents and that no single benchmark tests an agent across all its data surfaces under one policy.