Stripped Alignments: AI Safety Layers, Herd Immunity Thresholds, and Border Tower Coverage All Fail at the Same Structural Gap

Three unrelated Factum items reveal the identical failure mode. Heretic's release of Llama-3-8B weights with alignment layers removed directly parallels the Allegheny County MMR coverage drop to 91.8 percent—both cases show an engineered protective layer (model weights or population immunity) being locally stripped until the nominal safeguard no longer functions. The same pattern appears in the MIT TR border analysis: 138 deaths recorded inside the documented coverage radius of 587 surveillance towers since 2022. In each instance the system reports full coverage on paper while real-world exposure occurs precisely where the protective layer was thinned or absent. No single desk treated the AI de-alignment story as continuous with either the public-health threshold breach or the physical surveillance gap; all three are measurements of the same distance between declared protection and actual contact surface.