White House alleges Chinese entities stealing advanced AI technology
White House alleges Chinese AI technology theft; China denies accusations.
Event / Technology
US-China AI Competition
Chinese AI company DeepSeek announced DeepSeek-V4 with one million token context length, supporting multimodal content and Huawei chip compatibility. The announcement coincided with OpenAI's GPT-5.5 release and ongoing US-China tensions over AI technology theft accusations and market competition.
5 reported locations
Daily activity
0 articles0 articles
2026-08-18 to 2026-11-15 · UTC publication dates
Coverage begins 24 Apr, 12:07 UTC
Follow the documented timeline or explore the related reporting.
What happened
White House accuses Chinese entities of large-scale AI technology theft
The White House accused Chinese entities of conducting industrial-scale campaigns to clandestinely copy American artificial intelligence models. The accusation did not specifically name DeepSeek but was made in the context of intensified China-US AI competition.
DeepSeek's January 2025 AI release significantly impacted American tech markets
In January 2025, DeepSeek released R1, a conversational AI agent with exceptional reasoning capabilities that rivaled American competitors like Gemini, ChatGPT, and Claude at lower cost. The release shocked global markets and caused a decline in American technology stock values.
OpenAI releases GPT-5.5 model ahead of DeepSeek announcement
American company OpenAI unveiled GPT-5.5, presented as the most advanced AI model on the market. GPT-5.5 is a new generation of the model underlying ChatGPT, which is used by nearly one billion people globally. The announcement was made on the same day DeepSeek announced its new model.
Chinese foreign ministry rejects AI theft allegations as unfounded
Chinese foreign ministry spokesman Guo Jiakun rejected accusations of AI technology theft as "totally unfounded," characterizing them as a slanderous campaign against the achievements of China's AI industry.
DeepSeek launches DeepSeek-V4 AI model with one million token context length
Chinese AI company DeepSeek announced the launch of DeepSeek-V4, a new artificial intelligence model featuring ultra-long context capabilities supporting one million tokens. The model is available in two versions: DeepSeek-V4-Pro and DeepSeek-V4-Flash (more economical). The model can process the equivalent of several novels and answer detailed questions across the entire text without losing information. DeepSeek stated the long-context capability is among the best in the world while remaining faster and cheaper than comparable systems.
DeepSeek-V4 supports multimodal content generation and Huawei chip compatibility
DeepSeek-V4 is presented as a multimodal model capable of generating text, images, and videos. The company confirmed the model can function with the latest chips manufactured by Chinese technology company Huawei, marking an important development for China's AI independence given ongoing restrictions on access to advanced Nvidia chips from the United States.
Read the coverage
2 alternative perspectives
DeepSeek releases advanced V4 model as AI race intensifies
DeepSeek released advanced V4 model as OpenAI launched GPT-5.5, escalating US-China AI competition.