The EU AI Act's General Purpose AI Transparency Mandates are certainly a part of the whole, but the idea they're always a necessary component, in every single circumstance, has a low prior probability, maybe 20%. Consider a large AI model developed for internal fraud detection at a bank in Amsterdam, never deployed publicly; the need for full public transparency mandated by Article 53 becomes much less critical there, perhaps a 10% necessity rather than 90%. We need to update our view: this relationship is often conditional on the specific deployment context, not an absolute dependency.