Zuckerberg opposes the ban on Chinese AI models: openness as a strategy, not a threat
Meta CEO Mark Zuckerberg sharply criticized those in Washington lobbying for a complete ban on Chinese artificial intelligence models. In his view, such bans are not a solution to the problem but rather a sign of weakness. He emphasized that American companies need to focus on identifying their own bottlenecks and systemic barriers, rather than trying to artificially restrict competitors.
Kimi K3 as a Catalyst for Debate
The discussion intensified after the release of the Kimi K3 model from Beijing-based Moonshot AI. The developers position it as the world's first open model of the 3T class, with 2.8 trillion parameters, native vision, and a context window of 1 million tokens. By benchmarks, it trails only proprietary giants—Anthropic's Claude Fable 5 and OpenAI's GPT 5.6 Sol. The question of whether the Chinese lab used distillation techniques (training one model on the responses of another) has already been raised at the White House level. In July, Director of the Office of Science and Technology Policy Michael Kratsios directly accused Moonshot AI of this, while Treasury Secretary Scott Bessent threatened sanctions and inclusion on "blacklists" for "covert industrial distillation attacks."
Tech Industry Unites in Defense of Openness
Zuckerberg's stance is not isolated. The Linux Foundation, together with Cisco, Dell, IBM, Meta, Microsoft, and Nvidia, signed a collective letter in defense of open-weight models. They argue this is necessary to ensure transparency and security of AI systems. Nvidia CEO Jensen Huang also strongly opposed bans, stating that good Chinese open models should be used, and a campaign to block them could harm the United States itself. Meanwhile, the White House is concurrently developing a mechanism for preliminary review of the most powerful systems, suggesting developers voluntarily provide access to frontier models 30 days before release.
My expert commentary: The AI market is entering a new phase where geopolitics clashes with pragmatism. Banning Chinese models will not stop progress—it will only create parallel ecosystems. Open weights are becoming not just a development tool but a key element of global technological competition. Washington's attempts to impose restrictions look like a belated reaction to the fact that China has already become a full-fledged player in this race.