Goldman Sachs: Chinese AI developers may shift to paid weights licensing
Goldman Sachs predicts Chinese AI developers may transition from open-source to paid commercial licensing for model weights. Ronald Keung, the bank's head of Asia internet research, noted that as Chinese AI adoption surges, companies will seek to boost revenue by requiring third-party providers to buy commercial licenses to serve models on their own infrastructure. Currently, most Chinese models are distributed under permissive licenses like the MIT License, allowing free use and modification. Keung highlighted rapid domestic adoption and growing interest from global SMEs and larger companies. The shift could close the open-source revenue gap.
Key facts
- Goldman Sachs predicts Chinese AI developers may shift to paid weights commercial licensing.
- Ronald Keung is the Hong Kong-based analyst and head of Asia internet research at Goldman Sachs.
- Chinese AI models are currently distributed under open-source terms like the MIT License.
- Paid licensing would require third-party providers to buy commercial licenses for inference.
- Domestic adoption of Chinese AI is very fast.
- Global SMEs and larger companies are considering using Chinese models.
- The shift aims to boost revenue capture for Chinese AI developers.
- The MIT License originated from the Massachusetts Institute of Technology.
Entities
Institutions
- Goldman Sachs
- Massachusetts Institute of Technology
Locations
- Hong Kong
- China