Since launching ChatGPT two years ago, OpenAI CEO Sam Altman has emphasized that his company's success depends on three key factors: chips, data, and money. The more resources invested; the more powerful AI OpenAI can develop. However, Chinese startup DeepSeek is challenging this formula with an AI model just as capable—built at a fraction of the cost.

Figure 1. DeepSeek's Open AI: A Game-Changer and Challenge for Sam Altman.

This isn’t just about DeepSeek or China. Around the world, open-source AI models are emerging, threatening OpenAI and Anthropic’s business model, which relies on charging premium access fees. Unlike Google and Microsoft, which have other revenue streams, these companies face increasing competition from free or low-cost alternatives. Figure 1 shows DeepSeek's Open AI: A Game-Changer and Challenge for Sam Altman.

DeepSeek’s new model, R1, has been praised for its transparency—something critics say OpenAI has lacked since GPT-3. Founder Liang Wenfeng, a hedge fund entrepreneur, assembled a team of Ph.D. graduates and leveraged 10,000 Nvidia chips acquired before U.S. export restrictions. Despite using only, a fraction of the computing power of its Western counterparts, DeepSeek’s breakthrough signals a shift in China’s AI ambitions—and a potential shake-up of the global AI market.

Cracking OpenAI’s Business Model

Since ChatGPT’s launch, OpenAI CEO Sam Altman has built his company around three key resources: chips, data, and money. The idea is simple—more investment in these areas means more powerful AI models. However, DeepSeek, a rising AI firm from China, has developed a model just as capable as OpenAI’s latest iterations—at a fraction of the cost. This challenges the fundamental notion that AI advancement requires massive financial and computational resources, potentially disrupting OpenAI’s business model.

The Power of Open-Source AI

Unlike OpenAI, which keeps its models proprietary and charges for access, DeepSeek is embracing a more transparent and open-source approach. Open-source AI models are rapidly improving, making it easier for developers worldwide to build alternatives to ChatGPT and Claude (Anthropic’s AI). If freely available models can rival OpenAI’s, companies may opt for these cost-effective solutions instead of paying for access, eroding OpenAI’s market share.

China’s Strategic AI Push

DeepSeek is part of a larger movement within China to develop homegrown AI technologies amid increasing U.S. sanctions. The company was founded by hedge fund entrepreneur Liang Wenfeng, who assembled a team of Ph.D. researchers to create human-level AI. Despite facing restrictions on high-end chips, DeepSeek managed to train its AI using only 1/50th of the computing power used by companies like OpenAI, Google, and Microsoft—proving that efficient AI development is possible without Western resources.

A Global Shift in AI Competition

DeepSeek isn’t the only company embracing open-source AI—other startups worldwide are developing models that threaten OpenAI’s dominance. Companies like Meta and Mistral are also releasing powerful, freely available AI models, challenging OpenAI’s paid-access approach. If this trend continues, OpenAI and Anthropic may struggle to justify their high costs when equally powerful and transparent alternatives exist.

The Future of AI—Adapt or Be Disrupted

For OpenAI to maintain its lead, it may need to pivot toward more transparency, lower pricing, or additional revenue streams like cloud computing (similar to Microsoft and Google). Meanwhile, DeepSeek’s emergence signals that AI development no longer belongs exclusively to tech giants with billion-dollar budgets. The battle between closed-source AI firms and open-source challengers is just beginning—and it could reshape the industry forever.

Source:Business Standard

Cite this article:

Priyadharshini S (2025),” Why Chinese Tech Firm Deepseek's 'Open AI' Could Be a Major Threat to Sam Altman", AnaTechMaz, pp. 213