China's DeepSeek that 'shocked' America and American technology companies has an update that it says .... to OpenAI’s o3 and Google’s Gemini 2.5 Pro

Chinese AI startup DeepSeek has announced a significant upgrade to its R1 model, named DeepSeek-R1-0528, which claims to boast enhanced capabilities in mathematics, programming, and general logic while reducing AI "hallucinations." The announcement, made via a post on the AI model platform Hugging Face, positions the new model as a formidable competitor to industry leaders like OpenAI’s o3 and Google’s Gemini 2.5 Pro.

When Deepseek stunned America's technology industry

DeepSeek stunned the global tech community in January with the original R1, a model that rivaled the performance of systems developed by much larger U.S. companies, reportedly at a fraction of the cost. The debut disrupted the AI industry, raising questions about the massive global spending on computational resources and sparking a wave of AI releases from Chinese tech giants like Alibaba Group Holding Ltd. and Zhipu AI. Built for under $6 million using less-advanced Nvidia H800 chips, R1’s efficiency challenged assumptions about the necessity of massive computational investments, triggering a $1 trillion stock market drop, including Nvidia’s $589 billion loss. DeepSeek’s open-source model, praised for its cost-effectiveness, raised concerns about U.S. AI dominance, prompting calls for stricter export controls and urgent policy reassessments

American lawmakers, citing national security, swiftly reacted. Representatives Gottheimer and LaHood introduced the No DeepSeek on Government Devices Act, alleging Chinese Communist Party data theft risks. NASA, the Navy, and Congress banned DeepSeek on government devices, fearing espionage due to its China-based servers and lax data policies. Companies like Oracle blocked employee use, wary of intellectual property leaks.

New DeepSeek model works on same foundation

The upgraded DeepSeek-R1-0528 builds on this foundation, offering what the company claims is "greater depth of reasoning" and performance approaching that of the world’s top AI models. “Its overall performance is now approaching that of leading models,” DeepSeek stated, signaling its ambition to compete at the forefront of the global AI race.The rise of DeepSeek has also catapulted its founder,

Liang Wenfeng

, into the spotlight. Once a relatively unknown figure, Liang became a symbol of China’s growing prowess in AI innovation following R1’s debut. In February, he was invited by President Xi Jinping to a high-profile meeting alongside prominent entrepreneurs like Alibaba co-founder Jack Ma and Tencent’s Pony Ma, underscoring DeepSeek’s significance in China’s tech landscape.

What makes timing of DeepSeek announcement important

The timing of the DeepSeek-R1-0528 announcement was notable, coming just hours before Nvidia Corp., the Santa Clara-based leader in AI chips, released its latest financial report. Nvidia, whose shares took a hit following R1’s initial launch due to concerns about reduced AI infrastructure costs, has since recovered as global investment in AI data centers remains robust. Nvidia’s latest forecast reaffirmed strong demand for its chips, suggesting that DeepSeek’s cost-efficient approach has not slowed the broader AI investment boom.DeepSeek’s rapid progress highlights China’s accelerating role in the global AI arms race. Industry analysts note that the startup’s ability to deliver high-performing models at lower costs could reshape the economics of AI development, challenging the dominance of U.S. tech giants.