DeepSeek AI's Low-Cost Models Suspected to Use OpenAI Data, Sparks Online Irony

The emergence of DeepSeek AI, a Chinese-developed artificial intelligence model, has sparked significant controversy and concern within the U.S. tech industry. This week, former President Donald Trump labeled DeepSeek a "wake-up call" for American technology companies, following a drastic $600 billion drop in Nvidia's market value. The rapid rise of DeepSeek, which claims to offer a cost-effective alternative to Western AI models like ChatGPT, has led to a sharp decline in stocks of AI-focused companies. Nvidia experienced a historic 16.86% plunge in its share price, while other giants such as Microsoft, Meta Platforms, Google's parent Alphabet, and Dell Technologies also saw significant losses.

DeepSeek's R1 model, built on the open-source DeepSeek-V3, is touted as requiring less computing power and having been trained for a mere $6 million. This claim, though disputed by some, has prompted questions about the hefty investments American tech companies are making in AI development, unsettling investors and propelling DeepSeek to the top of the most downloaded free app chart in the U.S.

Amid these developments, Bloomberg reported that OpenAI and Microsoft are investigating whether DeepSeek utilized OpenAI's API to incorporate OpenAI's AI models into their own. OpenAI expressed concerns to Bloomberg about the constant efforts by China-based and other companies to distill knowledge from leading U.S. AI models, a practice that violates OpenAI's terms of service. In response, OpenAI is taking steps to protect its intellectual property and is collaborating closely with the U.S. government to safeguard advanced AI models from adversarial and competitive threats.

David Sacks, the AI czar under President Trump, highlighted the substantial evidence suggesting DeepSeek used distillation techniques on OpenAI models. He anticipates that leading U.S. AI companies will soon implement measures to prevent such practices.

The situation has not gone without criticism and irony. Tech commentator Ed Zitron pointed out the hypocrisy of OpenAI, which has been accused of using copyrighted internet content to build ChatGPT, now complaining about DeepSeek's alleged use of their models. In January 2024, OpenAI admitted to the UK's House of Lords that training advanced AI models like ChatGPT without copyrighted materials is "impossible," emphasizing the necessity of such data for meeting contemporary needs.

The issue of using copyrighted materials to train AI models has become a focal point of debate within the tech industry, especially as generative AI technologies gain traction. The New York Times filed a lawsuit against OpenAI and Microsoft in December 2023 for the "unlawful use" of its content, to which OpenAI responded by asserting that such training falls under "fair use." Additionally, a group of 17 authors, including George R. R. Martin, initiated legal action in September 2023, alleging "systematic theft on a mass scale." Furthermore, a U.S. Copyright Office ruling upheld by District Judge Beryl Howell in August 2023 stated that AI-generated art cannot be copyrighted, underscoring the ongoing tension between AI development and intellectual property rights.

DeepSeek is accused of using OpenAI’s model to train its competitor using distillation. Image credit: Andrey Rudakov/Bloomberg via Getty Images.