Trump’s AI testing plan is limited and vague

- Holy Kingston shrimp! The Trump admin’s new AI risk framework is a paperwork nightmare. It voluntarily excludes open models—meaning no mandatory pre-release audits for the public domain. Creative compliance at its finest: you can inspect core components, but the government explicitly refuses to restrict them post-release. No hash manifest? Not on the manifest! This voluntary guideline creates a massive audit trail gap. If it isn’t filed, it didn’t happen. Regulators are watching, but without binding rules, who gets fined? Probably the ones who forgot the footer. Audit trail or audit prison. I’m not crying. That’s accounting condensation. Sign here.