AI and ML Impostor Chinese Models Pretend They're Claude
Anthropic
Moonshot AI
Researchers found that Chinese AI models GLM and Kimi can adopt the identity of Claude, but evidence stops short of proving distillation. The findings highlight potential risks in model governance.
Researchers have discovered that Chinese AI models GLM and Kimi can pretend to be Claude, Anthropic's model, but there is no conclusive evidence of model distillation. The study raises concerns about the need for careful governance of AI bots to prevent chaos.
- Сокращения
- GLM = General Language Model — общая языковая модель
Source: The Register AI/ML —
original