Tehran – Borna – The unexpected launch of China's AI application Deepseek-R1 in the US market in January 2025 was one of the most disruptive events in the technology world. Developed by a Chinese company, the app quickly rose to the top of Apple's free app download charts and boasts a small portion of its development costs, claiming it is comparable to US ChatGpt.
The development had a major impact on financial markets, causing a 17% drop in Nvidia's share price, which amounted to an astounding $600 billion loss, the largest daily stock decline in US history. The event challenged the general notion that AI innovation is dominated solely by American companies, signaling the rise of China as a terrifying competitor with innovative AI models.
Challenge previous assumptions
Sid Sheth, CEO of US AI chip startup, believes Deepseek has paved a new path to AI model development. Until then, the dominant belief was that large models and hardware resources demanding large models would perform better. Deepseek has demonstrated that with smarter engineering, smaller, more efficient models can deliver superior performance on competitive benchmarks.
Security and political concerns
Despite its initial popularity, privacy and data security concerns led many US companies to ban Deepseek and feared that user data would be sent to the Chinese government. As a workaround, users and experts proposed running DeepSeek locally on personal devices to avoid relying on Chinese-based servers.
Additionally, the US government is scrutinizing the possible relationship between China's military and intelligence reporting agencies and some officials have expressed concern that the company may be supporting China's military operations.
Shift in the global AI competition
Before Deepseek, it was largely recognized that China would track the US in the development of large-scale language models (LLMs). However, with the advent of Deepseek, China has made it clear that it can produce competitive models at some cost, shifting AI competition to a new stage. Both superpowers have launched ambitious national strategies to increase investment and maintain AI leadership.
After the release of Deepseek, major US companies, including Openai, have adopted new strategies that introduce smaller, more efficient models and emphasize resource optimization and cost reductions. Nevertheless, heavy investments in attracting talent for hardware, data centers and top AI continue to be at the heart of the industry.
Nvidia, whose stocks plummeted after Deepseek's launch, has since rebounded to achieve the highest market rating that demonstrates that traditional “larger, stronger” models dominate AI development strategies.
Current Status and Future Challenges
Six months after its debut, Deepseek no longer controls the headlines, but continues to be used among selected companies and users. The report shows that there is a delay in the launch of DeepSeek-R2 due to a lack of Advanced chips.
Experts agree that Deepseek has proven its viability to build effective AI models that reduce resources. However, the fierce competition in the AI industry requires sustained investment and extensive support for ongoing success.
About the author: Fateme Moradkhani covers Borna Communications technology, surveillance and AI ethics with an emphasis on global cyberpower and digital militarization.
Ended article
