ChatGPT Maker Suspects China’s Dirt Cheap DeepSeek AI Models Were Built Using OpenAI Data — and the Irony Is Not Lost on the Internet

The emergence of DeepSeek AI, a Chinese company offering ultra-cheap AI models, has raised suspicions among the developers of ChatGPT, who believe that DeepSeek may have used OpenAI's data to train its models. This development has caused significant turbulence in the U.S. tech industry, with former President Donald Trump labeling DeepSeek as a "wake-up call" following a massive $600 billion drop in Nvidia's market value.

The impact of DeepSeek's entry into the market was immediate and severe, sending stocks of AI-heavy companies into a downward spiral. Nvidia, a dominant player in the GPU market crucial for AI operations, experienced a historic 16.86% plunge in its stock price. Other tech giants like Microsoft, Meta Platforms, and Google's parent company Alphabet saw declines ranging from 2.1% to 4.2%, while Dell Technologies, a key AI server manufacturer, dropped by 8.7%.

DeepSeek's R1 model, built on the open-source DeepSeek-V3, claims to offer a significantly more affordable alternative to Western AI models, requiring less computing power and reportedly trained for just $6 million. Despite some skepticism regarding these claims, DeepSeek's model has challenged the hefty investments American tech companies are making in AI, causing investor unease. The app's popularity surged, propelling it to the top of the U.S. free app download charts amid discussions about its effectiveness.

In response to these developments, Bloomberg reported that OpenAI and Microsoft are investigating whether DeepSeek utilized OpenAI's API to integrate OpenAI's AI models into their own. OpenAI expressed concerns about Chinese companies and others attempting to distill data from leading U.S. AI models, a practice that violates their terms of service. OpenAI emphasized its commitment to protecting its intellectual property and hinted at potential collaboration with the U.S. government to safeguard advanced AI models from competitors and adversaries.

David Sacks, the AI czar appointed by President Trump, suggested on Fox News that there is "substantial evidence" that DeepSeek employed distillation techniques to extract knowledge from OpenAI's models. He anticipated that U.S. AI companies would take steps to prevent such practices in the coming months.

Amidst these accusations, critics have pointed out the irony, given OpenAI's own history of using copyrighted material to train ChatGPT. Tech commentator Ed Zitron highlighted this hypocrisy, noting OpenAI's complaints about DeepSeek potentially training on ChatGPT outputs while having built its own technology on internet data.

In early 2024, OpenAI acknowledged the necessity of using copyrighted materials to train large language models like ChatGPT, stating in a submission to the UK's House of Lords that excluding such materials would render modern AI systems inadequate for today's needs. This stance comes amidst ongoing legal battles over AI training practices, including a lawsuit from The New York Times against OpenAI and Microsoft for the alleged "unlawful use" of its content. OpenAI has defended its practices as "fair use" and maintained its support for journalism.

The legal landscape surrounding AI and copyright continues to evolve, with significant cases such as the lawsuit by 17 authors, including George R. R. Martin, against OpenAI for "systematic theft on a mass scale," and a ruling by District Judge Beryl Howell affirming the U.S. Copyright Office's stance that AI-generated art cannot be copyrighted due to the essential human element required for copyright protection.