China's DeepSeek AI Models: Built with OpenAI Data? Internet Sees Irony

The emergence of DeepSeek AI, a Chinese-developed model touted for its affordability, has sparked controversy and concern within the U.S. tech industry. Donald Trump labeled DeepSeek as a "wake-up call" following a significant $600 billion drop in Nvidia's market value, reflecting broader market anxieties about AI investments.

DeepSeek's R1 model, built on the open-source DeepSeek-V3, claims to offer a cost-effective alternative to Western AI models like ChatGPT, requiring considerably less computing power and reportedly trained on a budget of just $6 million. This has led to a sharp decline in stock values for major AI-related companies, with Nvidia experiencing a historic 16.86% drop, while Microsoft, Meta Platforms, Google's Alphabet, and Dell Technologies also saw significant losses.

The rapid rise of DeepSeek, which topped the U.S. free app download charts, has prompted OpenAI and Microsoft to investigate whether DeepSeek used OpenAI's API to train its models, potentially violating OpenAI's terms of service through a technique known as distillation. OpenAI emphasized the importance of protecting its intellectual property and hinted at future collaborations with the U.S. government to safeguard advanced AI technologies.

David Sacks, President Trump's AI czar, suggested that leading U.S. AI companies would take steps to prevent such distillation practices. The situation has drawn ironic commentary, with critics like Ed Zitron highlighting OpenAI's own history of using copyrighted material to train ChatGPT, as admitted in a submission to the UK's House of Lords.

OpenAI has argued that training AI models on copyrighted materials is necessary and falls under "fair use," despite facing lawsuits from entities like The New York Times and a group of authors led by George R. R. Martin. These legal challenges underscore the ongoing debate over the use of copyrighted content in AI development, further complicated by a U.S. Copyright Office ruling that AI-generated art cannot be copyrighted due to the lack of human creative input.

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