Chinese AI Sparks OpenAI Data Theft Allegations

OpenAI suspects that China's DeepSeek AI models, significantly cheaper than Western counterparts, may have been trained using OpenAI data. This revelation, following Nvidia's massive stock plunge, has sent shockwaves through the AI industry.

Donald Trump called DeepSeek a "wake-up call" for the US tech sector. The release of DeepSeek triggered a sharp decline in AI-related stocks, with Nvidia experiencing its largest-ever single-day loss (16.86%). Microsoft, Meta, Alphabet, and Dell also suffered significant drops.

DeepSeek boasts its R1 model as a cost-effective alternative to ChatGPT, trained using the open-source DeepSeek-V3 for an estimated $6 million. While this claim is debated, it has raised concerns about the billions invested by American tech firms in AI, impacting investor confidence. DeepSeek's app also surged in popularity in the US.

OpenAI and Microsoft are investigating whether DeepSeek violated OpenAI's terms of service by using its API for model distillation – a technique to train AI models using data from larger ones. OpenAI confirmed its awareness of such attempts by Chinese and other companies and emphasized its efforts to protect its intellectual property (IP) and collaborate with the US government.

David Sacks, President Trump's AI advisor, stated there's substantial evidence of DeepSeek's use of OpenAI models, predicting countermeasures from leading AI companies.

The situation highlights the irony of OpenAI's accusations, given previous controversies surrounding its own use of copyrighted internet data for ChatGPT. Ed Zitron aptly pointed out this hypocrisy on Twitter. OpenAI previously argued that creating AI tools like ChatGPT without copyrighted material is impossible, a stance reiterated in a submission to the UK's House of Lords. This argument is further complicated by ongoing lawsuits, including one from the New York Times alleging unlawful use of its work, and another from 17 authors claiming "systematic theft." The legal landscape surrounding AI training data and copyright remains highly contested, particularly in light of a 2018 US Copyright Office ruling that AI-generated art is not copyrightable.