At the 2026 World Internet Conference Wuzhen Summit on November 3, researchers from DeepSeek, Alibaba's Qwen team and MiniMax shared a stage in a session titled the Wuzhen Dialogue on Artificial Intelligence, moderated by Wang Jian, academician of the Chinese Academy of Engineering and director of Zhejiang Lab, according to The Paper. Panelists included DeepSeek research engineer Xi Gao, Qwen LLM project lead Dayiheng Liu, MiniMax chief architect and head of open platform Wen Feng, and John Higgins, chair of the International AI Governance Association. Gao said DeepSeek open-sources to reach artificial general intelligence sooner, viewing time rather than other companies as the real competitor. Liu said Qwen has open-sourced more than 460 models since releasing its first generation three years ago, with cumulative global downloads above 4 billion and more than 330,000 derivative models. Feng said MiniMax aims to contribute differentiated value by releasing models across modalities and sharing engineering experience. On risk, Liu identified irrevocability and the lack of external guardrails as weaknesses specific to open models, calling for laws, industry norms and supporting infrastructure, while noting openness also helps safety. Feng said open models are often the more controllable choice for industry. Gao acknowledged open source may amplify AI safety risks and said DeepSeek has a responsibility to keep them from spreading and is researching solutions.