DeepSeek returns after year-long silence, researcher warns of AI’s threat to human jobs

China’s artificial intelligence (AI) developer DeepSeek made a rare public appearance in nearly a year, with one of its senior researchers voicing concern over the long-term impact of AI on humanity. Speaking at the World Internet Conference in Wuzhen, Zhejiang province, Chen Deli said he was “extremely positive about the technology” but “view[ed] the impact it could have on society negatively.”

At the government-organised event, Chen shared the stage with the chief executives of five other Chinese AI firms, including Unitree and BrainCo. Collectively dubbed China’s “six little dragons” of AI, the companies are seen as emerging leaders in the country’s fast-growing artificial intelligence ecosystem.

When asked about DeepSeek’s global success and its open-source approach, Chen struck a cautious tone. He said that while AI could significantly aid humanity in the short term, the technology poses serious employment risks in the coming decade.

“AI could be a great aid to humans as it improves over the short term, but it could threaten job losses in 5–10 years as it becomes good enough to take over some of the work humans perform,” he said.
“In the next 10–20 years, AI could take over the rest of work (humans perform) and society could face a massive challenge, so at the time tech companies need to take the role of ‘defender’.”

Chen added that while he remains optimistic about AI’s potential, he is concerned about its broader social effects.

DeepSeek rose to global prominence in January 2025, after releasing a low-cost AI model that outperformed leading U.S. systems. Its breakthrough drew international attention, positioning the company as a symbol of China’s technological progress amid deepening U.S.–China tensions over advanced tech.

Since then, the company has largely stayed out of public view. Its founder and CEO Liang Wenfeng last appeared publicly in February, when he met Chinese President Xi Jinping during a televised gathering of entrepreneurs. DeepSeek has since skipped major domestic tech conferences, maintaining a low profile even as its influence in China’s AI race grows.

In September, DeepSeek unveiled an upgrade to its V3 model, calling it an “experimental version” with more efficient training and stronger long-text processing. The company has also aligned itself with the government’s push to strengthen domestic chip capabilities. Its AI models are now compatible with Chinese-made hardware developed by Huawei and Cambricon, both of which have become key players in China’s semiconductor ecosystem.

In August, DeepSeek’s announcement of an AI model optimized for Chinese chips sparked a rally in domestic semiconductor stocks — a sign of investor confidence in Beijing’s bid to reduce reliance on U.S. technology.

While DeepSeek has not released another major model since January, its moves continue to shape China’s AI ambitions and the global conversation around open-source innovation, national resilience, and the ethics of artificial intelligence.