At least one of the big AI brains at the Chinese chatbot leader DeepSeek has announced that he’s worried—like many U.S. AI luminaries before him—that AI will hurt the world.
But in China, unlike the U.S., the state can and does regulate tech with a heavy hand, and in that context such a sentiment comes across as slightly more substantive than when, say, OpenAI’s Sam Altman hypes up his future investment by saying things like, “A.I. will probably most likely lead to the end of the world, but in the meantime, there’ll be great companies.”
For context, at the start of this year, DeepSeek, the little-known AI wing of a hedge fund based in Hangzhou, Zhejiang Province exploded onto the scene, having suddenly come into possession of the holy grail of the tech world: its own cutting-edge large language model, DeepSeek-R1.
Following its breakout success, the DeepSeek team has been pretty tight-lipped for most of the year, but DeepSeek senior researcher Chen Deli has finally broken the company’s silence on whether or not what they are doing is good or bad for society. According to Chen: quite possibly bad!
“I’m extremely positive about the technology but I view the impact it could have on society negatively,” Chen said onstage in Wuzhen at the government-organized World Internet Conference.
Back in January, Chen was in a better mood, tweeting that he was “beyond words right now,” and that the astronomical benchmark ranking of his company’s low-cost model “feels like a dream.” That was January. Then in February, Deepseek’s Founder Liang Wenfeng met with President Xi Jinping at a televised symposium. After that, as Reuters notes, no one from the company has addressed the public, until Chen’s depressing appearance at the World Internet Conference.
According to Chen, AI will threaten jobs in the next decade as it becomes good enough to get people replaced, but AI firms have a job to do when it comes to staving off disaster for humanity. “Society could face a massive challenge,” Chen said, “so at the time tech companies need to take the role of ‘defender’.”
Realistically, U.S. companies are probably not the companies Chen has in mind as “defender.” His company has close ties to other Chinese companies, many of which have scrambled to integrate DeepSeek AI products into their own ever since the R1 model was released earlier this year, apparently part of the increasingly DeepSeek-centric plan for “algorithmic sovereignty”—cutting-edge achievements in AI without foreign involvement, basically. A little over a month ago, DeepSeek released what it called an “experimental” version of its LLM, a version notable not so much for its efficiency gains or new capabilities, as for its creation of an alternative to Nvidia’s CUDA API, and its support for Chinese GPUs.
A little over a week ago, Xi said at the 2025 Asia-Pacific Economic Cooperation forum that there should be a global body that governs AI, and that such an international governmental organization would be in the interest of making AI a “public good for the international community.”
Now a representative from DeepSeek itself is making statements about AI being a potential hazard for society that the tech world must find a unified approach to mitigating.
Yes, when an AI billionaire from the U.S. says this sort of thing, it’s probably just hype. Chen’s statement, on the other hand, sounds less like tech business swagger and more like the literal party line.
