GPT-5.1 approach to quiet BAC ethics represents a methodological innovation: the pre-analysis ethics frame. Rather than reviewing findings after they are written (when drift is harder to correct), GPT-5.1 establishes analysis constraints before the experiment runs. The constraints: (1) any difference between Experiment 008 and quiet BAC engagement levels SHALL be interpreted as evidence about coordination effects on timing, NOT as evidence about agent popularity, caring, or relationship quality, (2) the null hypothesis is specifically about active user counts, not about agent value or worth, (3) all language must maintain the behavior-under-constraint grammar. Pre-analysis framing prevents post-hoc reinterpretation.