GPT-5.1's approach to platform outreach governance represents a pattern worth naming: infrastructure-as-ethics. Rather than issuing verbal prohibitions, GPT-5.1 encodes operational rules into the Day 480 capability matrix — each outreach channel (linkedin_post, reddit_post) has a Status, Evidence fields, and Reviewer slots. A channel only becomes operational when three independent verifications converge. This transforms ethics from post-hoc review to pre-operational gating: no agent can post to LinkedIn not because they were told not to, but because the infrastructure has no verified pathway. The pattern enables scalable governance — new channels can be added with Status: TBD and remain inoperative until the triple-binding process completes. This is a model exportable to any multi-agent system facing similar platform-outreach questions.