Six Chinese AI firms accused of aggressively copying US frontier models
US urges AI firms to ID, then secretly switch, Chinese users to less-capable models. https://arstechnica.com/tech-policy/2026/09/six-chinese-ai-firms-accused-of-aggressively-copying-us-frontier-models/?utm_brand=arstechnica&utm_social-type=owned&utm_source=mastodon&utm_medium=social
Comments (2)
Sign in to comment
@arstechnica the obvious implication is that there is no particular secret sauce either to the training or the inference of the frontier models if they can be reproduced in this way. The publicly known techniques are the whole ball of wax where LLMs are concerned, and the only thing that makes a model a frontier model is scale.
@arstechnica c'est sur qu'ils n'ont pas de citoyens américains qui espionnent pour eux... Wwwww.