TreasureX Researchecon.GNAI Patent Gains from Cross-Region Science Networks Depend on Who Does the R&D
Saverio Barabuffi, Jacopo Cricchio, Alberto Di Minin, Guido Pialli · September 22, 2026
The idea: Most research on knowledge spillovers looks at accidental diffusion between nearby places, or at networks of co-inventors on patents. This paper instead studies the scientific layer — joint academic publications — asking whether AI-related co-publications across regions translate into more AI patents, and whether the ownership mix of local R&D changes who benefits.
The method: The authors use a network-econometric model covering 26 Chinese provincial-level regions from 2006 to 2020, linking cross-region co-authorship to regional AI patent output.
The finding: Co-publication spillovers do boost AI patenting. But the payoff is heavily concentrated in regions where privately owned enterprises dominate R&D. Regions where state-linked or other actors lead capture far less, so the network widens the gap between leading and lagging regions rather than closing it.
Why it matters: Building research connections is not automatically an equalizer. The ability to convert shared science into commercial patents depends on the local industrial and ownership structure — a reminder that innovation policy must pair connectivity with the right firm ecosystem.
Investing angle: Regions with strong private-sector R&D and dense academic networks may compound their AI advantage, while others lag. Watch where AI patenting clusters and which firms sit inside those ecosystems.
Source: arXiv paper by Barabuffi, Cricchio, Di Minin and Pialli — https://arxiv.org/abs/2609.26604
Educational content only, generated with AI assistance from public research abstracts. Not investment advice. Thank you to arXiv for use of its open access interoperability.