DeepSeek and Global AI: A Race for Supremacy, Not Ethics

As Silicon Valley panics and sets up a "war room" to understand how the Chinese startup DeepSeek managed to outperform established AI models like ChatGPT, Llama, and Gemini, the rise of this technology does not necessarily represent a democratic or ethical advancement in artificial intelligence. While Chinese AI promises accessibility and open-source technology, the geopolitical realities and strategic interests behind its development reveal that the battle remains among giants such as OpenAI, Google, Meta, Microsoft, Anthropic, and now DeepSeek - all competing for technological dominance with little genuine concern for the political and social impacts of this technology.

Founded in 2023 by Liang Wenfeng, with significant contributions from Luo Fuli, a young woman AI prodigy, DeepSeek has achieved in less than two years what took OpenAI nearly five years to accomplish. Its latest model, DeepSeek R1, has surpassed ChatGPT in downloads and benchmarks, proving more efficient in solving mathematical problems, programming, and logical reasoning. The major advantage? An operational cost up to 95% lower than its top competitors.

With a training cost of $6 million, a fraction of the estimated $100 million needed for ChatGPT-4, and relying on less advanced chips, DeepSeek challenges the narrative that only the United States possesses the most sophisticated AI technology. This has led analysts to compare its rise to the “Sputnik moment” of artificial intelligence: a reference to the impact of the Soviet Union’s launch of the first satellite, which redefined the space race and sent shockwaves through the United States.

The market reaction was rapid. Nvidia, the largest supplier of AI chips, saw its stock plummet by 12%, resulting in a $300 billion market cap loss. Other tech giants like Meta, Microsoft, and Google also experienced significant drops in stock prices. Meanwhile, DeepSeek became the most downloaded app on the U.S. App Store, overtaking ChatGPT.

Accessible Yet Controlled: The Limits of DeepSeek's ‘Open’ AI

DeepSeek advocates for an open-source approach, which, in theory, would allow greater access to advanced AI models, bypassing the licensing barriers imposed by Western companies. However, this so-called "democratization" does not equate to neutrality or transparency. The model strictly adheres to the ideological directives of the Chinese Communist Party, censoring sensitive topics such as the 1989 Tiananmen Square Massacre, Taiwan’s independence, and Hong Kong’s autonomy. DeepSeek also reinforces Beijing’s official stance on issues such as the Russia-Ukraine war and the Gaza conflict, aligning itself with Chinese diplomatic narratives.

While OpenAI and Google take a cautious approach to political topics, often avoiding direct involvement, DeepSeek’s AI openly aligns with the Chinese government, repeating its official stance. This raises a key concern: even though China has bypassed U.S. sanctions and built a strong AI model, its technology is not a neutral source of information. The Great Firewall is no longer limited to China. It has expanded under the pretense of technological progress.

DeepSeek’s rapid rise is not happening in a vacuum. It is part of a larger geopolitical struggle between the U.S. and China for AI dominance, where both sides are racing to secure technological and economic hegemony. The United States has restricted the export of advanced AI chips to China since 2023, betting that this would slow down Chinese AI development. Yet, DeepSeek has proven that it is possible to build highly efficient AI models even with limited hardware, raising serious doubts about the effectiveness of these sanctions.

Yet, the rush for AI supremacy overshadows a more fundamental issue: the absence of a serious debate on AI governance and its societal impact. Who controls this technology? How can we prevent AI systems from being used to reinforce inequalities, mass surveillance, and misinformation? How do we ensure that AI serves society as a whole, rather than just powerful corporate and political interests?

Today, AI is developed and controlled by a small group of corporations and governments, prioritizing their own interests over the need for robust global regulation. The DeepSeek case exposes this reality: while its open-source model might be seen as an accessibility breakthrough, its state-aligned censorship and ideological filtering reveal that AI remains in the hands of the most powerful.

The idea that AI can be a democratizing force collapses when examined in a global context. This is not a battle for a more ethical or inclusive AI. It is a fight for who gets to set the rules.

In the end, we are witnessing a battle among lions, where the impact of AI on society is secondary. As governments and corporations compete for technological supremacy, the world still lacks an effective regulatory framework prioritizing transparency, human rights, and safeguards against AI-related harms.

If there is one key lesson from DeepSeek’s rapid ascent, it is not just about its technological breakthrough or its disruptive impact on Silicon Valley. The stark realization is that without global debate and strong AI governance policies, artificial intelligence will remain a tool of power, wielded by the few and inevitably used to reinforce the interests of the most dominant players.