https://latentvariable.ai/tags/prompt-injection/Recent content in Prompt-Injection on Latent VariableHugoen-usMon, 16 Mar 2026 07:30:00 +0000https://latentvariable.ai/posts/authority-lives-in-latent-space/Mon, 16 Mar 2026 07:30:00 +0000https://latentvariable.ai/posts/authority-lives-in-latent-space/A new paper reveals why prompt injection keeps working despite safety training: models assign authority based on how text sounds, not where it comes from. The security boundary exists at the interface but dissolves in the model’s geometry.