The U.S. Center for AI Standards and Innovation (CAISI) scrubbed its public website of nearly all documentation regarding safety agreements with Google, Microsoft, and xAI this week. What was once a public record of how these companies would vet their frontier models has vanished into a classified, proprietary void. This cognitive enclosure moves the oversight of our most powerful technologies out of the public square and into the hands of a small elite. According to officials familiar with the move, the scrubbing of these records effectively hides the 'moat' of safety protocols that are being built to protect corporate interests rather than the public common. This occurs as 'vibe coding'—the use of amateur AI tools to build software—has already leaked over 380,000 sensitive corporate assets online. Without public oversight, the digital world becomes a gated hub where the rules are written in secret by the very people they are meant to govern. Politics is the moat that guards the treasures of our civilization, and transparency is the gate. When that gate is locked from the inside by tech giants and their federal partners, the treasures of our faith, family, and free speech are no longer secure.