Chinese artificial-intelligence firms have largely stayed silent in response to high-profile warnings from leading U.S. technology companies about the risks of AI development, opting instead to focus on commercial application and regulatory compliance.
Z.ai, Moonshot, MiniMax, Alibaba and Tencent did not comment when contacted about recent alerts from Silicon Valley. Their restraint contrasts sharply with a rare public display of industry unity among U.S. executives earlier this month, when OpenAI CEO Sam Altman, Elon Musk and Anthropic CEO Dario Amodei called for a slowdown in AI development citing uncontrollable risks. Nvidia CEO Jensen Huang countered by arguing that speed and safety can coexist and urging responsible development practices.
In his latest essay, Amodei urged U.S. companies to maintain a technological lead over China, echoing language from a paper Anthropic published in May. Chinese officials quickly dismissed the warnings. The foreign ministry called them "fear-mongering," while English-language state media op-eds described the rhetoric as a "Cold War playbook" and invoked "Dr. Frankenstein."
"Once Chinese companies have created superior AI, we see U.S. companies issuing AI warnings. I don't think this is a coincidence," said Renjie Guo, founder and CEO of JoyIn. "From a philosophical perspective, I expect AI is only as dangerous as its developers."
Beijing moved simultaneously to reinforce its own regulatory posture. During its annual cybersecurity week, the government released the third edition of an "AI Safety Governance Framework," a bilingual document outlining guidelines for labeling AI-generated content and developing rapid risk-detection systems.
Chinese firms have taken a fundamentally different approach to AI development than their American counterparts. Constrained by U.S. restrictions on access to advanced semiconductors, Chinese companies have concentrated on commercialization rather than the pursuit of artificial general intelligence. Beijing began regulating AI outputs early, keeping domestic alternatives to ChatGPT off the public market for months after the platform launched before greenlighting them in the summer of 2023.
By early 2026, Chinese AI chatbots were competing aggressively for users with large promotional campaigns around the Lunar New Year. Several domestic models vied with Anthropic's Claude and OpenAI's ChatGPT on performance benchmarks, often at significantly lower cost. The cheaper open-source variants, particularly from DeepSeek and Alibaba's Qwen, have attracted users in the United States and elsewhere.
The content filtering remains consistent regardless. Asked about June 4, 1989, DeepSeek declined to respond, while it provided detailed information about the September 11, 2001, attacks.
China's cybersecurity regulator maintains formal approval processes for generative AI services likely to influence public opinion and publishes official lists of registered models. President Xi Jinping stressed the need to "ensure that AI is secure and controllable" during the July launch of the World Artificial Intelligence Cooperation Organization, and later called for deeper AI cooperation among BRICS economies at the group's summit in India.
Chinese companies emphasize revenue generation over intelligence rankings. "We are not looking for the most impressive AI models, but trying to unlock AI potential," said Alex Lu of LSY Consulting, noting that many firms rely on harness technology and reinforcement learning rather than the most advanced architectures.
Ray Von, CEO of Tencent-backed OpenPie, said his company primarily uses older, smaller versions of Qwen and DeepSeek models. "The vast majority of enterprises see a lot of productivity increases but they haven't seen anything generate profits or reduce costs immediately," he said. "Right now, the publicity is diverting a lot of attention."
