TY  - RPRT
TI  - Evaluating Epistemic Guardrails in AI Reading Assistants: A Behavioral Audit of a Minimal Prototype
AU  - Matthew Christian Agustin
PY  - 2026
UR  - https://arxiv.org/abs/2604.27275
ID  - 2604.27275
ER  -