Amid intensifying competition in artificial intelligence between the United States and China, the emergence of the DeepSeek-R1 model has created significant ripples across the technology sector, capital markets, and policy circles. This presentation analyses the legal and policy landscape surrounding DeepSeek, drawing upon its key technical features—including reinforcement learning, a mixture-of-experts architecture, a multi-head latent attention mechanism, knowledge distillation, and an open-source approach.

Firstly, DeepSeek’s success raises critical questions about the efficacy of the United States’ increasingly stringent export control measures on chips and semiconductors—components essential for training large language models. Secondly, much like Chinese technology giants such as TikTok and Huawei, DeepSeek has come under information security scrutiny in the United States and other jurisdictions. Thirdly, despite its domain-specific capabilities and competitive API pricing, DeepSeek has faced criticism for generating outputs laden with political and ideological biases, thereby igniting debates over free speech and censorship. Furthermore, this presentation examines intellectual property and contractual concerns stemming from the knowledge distillation technique employed by DeepSeek. In conclusion, the presenters are of the view that geopolitical considerations will continue to exert a profound influence on the legal challenges and possible solutions related to the DeepSeek models.