DeepSeek a year on: how a little-known Chinese start-up sparked a global AI arms race

One year ago, a little-known Chinese start-up called DeepSeek burst onto the scene with a new artificial intelligence model that challenged assumptions about China’s ability to innovate under US technology curbs.In what became known as the “DeepSeek moment”, the Hangzhou-based firm kicked off what some likened as a modern-day “Sputnik moment” for China’s AI ambitions.Just as the Soviet Union’s 1957 launch of Sputnik, the world’s first artificial satellite, shocked the US and triggered the space race, DeepSeek’s open-source R1 reasoning model released on January 20, 2025, was – in the words of US President Donald Trump – a “wake-up call” for the US tech industry.“DeepSeek will always have a unique place [in Chinese AI development] as the lab that changed the tone of this co-opetition between the US and China,” said Kevin Xu, founder of technology hedge fund Interconnected Capital.Its emergence, he said, suggested that China was closer to the US in advanced AI development than many had previously believed.Industry experts said DeepSeek’s impact for China goes beyond technology. Photo: DigiTimesThe breakthrough by the Hangzhou-based lab in training advanced AI systems with inferior chips on a shoestring budget in the face of sweeping US export controls underscored its intention to help China overcome the export ban, said Chelsey Tam, senior equity analyst at investment consultancy Morningstar.