Woofun AI reports that former OpenAI researcher Will Depue sparked controversy by alleging China has accessed all GitHub repositories, Slack records, and internal documents from OpenAI and Anthropic. Depue suggested that plagiarized architectures in Chinese open-source models are plausible, citing conversations with cybersecurity experts who doubt the efficacy of current lab defenses.

The claim drew sharp criticism from the AI community. Scholar Nathan Lambert dismissed the narrative as self-serving, emphasizing that technology transfer relies on talent mobility and engineering resources rather than chat logs.

Meanwhile, researcher Elie Bakouch ridiculed the sentiment, noting that US labs also benefit from Chinese open-source models like DeepSeek and Kimi.