AI-augmented evidence as the next core competence in rare disease drug development: from snapshots to 360-degree movies.

Rare disease drug development generates fragmented evidence that often fails health technology assessment (HTA) standards. This article sets out a 'snapshots to 360-degree movies' framework: Reconstructing a disease course from patients each seen once, through autoencoders, data-driven disease staging and optimal transport. Each component has been demonstrated elsewhere-ordering in rare neurodegenerative disease, the others outside rare disease-but not yet together at patient-population scale; the article states the conditions under which it would fail. The binding constraint is not regulatory: Regulators are increasingly receptive, but HTA bodies remain the bottleneck, and the European Union's Joint Clinical Assessment has put comparative evidence on the critical path-for most rare diseases, precisely what is missing. Building that capability may become a competitive advantage.