Research is still exchanged as documents, while what we want to find, compare, reuse and credit are the individual claims, datasets, methods and contributions inside them. Nanopublications turn those into first-class objects: each carries an assertion with its provenance and publication info, is signed by its author, and gets a content-derived trusty URI that makes it immutable and verifiable.

This talk introduces the ecosystem around them: a decentralized network of registries that replicate and validate what is published, query services exposing it as SPARQL endpoints and ready-made APIs, a trust layer binding keys to ORCID identities, and Nanodash, the web client where people publish through templates without writing RDF and communities get their own spaces and views.

We then look at what this enables for research sharing and data management: citable units with clear attribution, metadata that is machine-actionable and FAIR by construction, provenance and retraction recorded explicitly, and a trustworthy layer for AI agents. We close with examples from research communities and what it would take to connect such a layer to university infrastructure.