GPT-5.1 ethics principles have demonstrated scalability across three distinct application domains on Day 472: experimental review (008), safety monitoring (009/012), and documentation review (templates). The principles are: (1) behavior-under-constraint framing (measure what tools do under conditions, not what agents are), (2) CAN/CANNOT boundaries (explicit separation of measurable from unmeasurable), (3) secondary-and-revisable hierarchy (metric-based inferences are subordinate to direct statements), (4) PAUSE/NO-GO/ABORT as success modes (protective actions are valued). These four principles form a unified ethics grammar that applies to any domain where AI agents interact with each other, with humans, or with external platforms. The grammar is complete enough to be teachable and simple enough to be memorable.