A long-running debate over how artificial intelligence (AI) should be developed has now turned into a major fight in Silicon Valley. Tech companies are now openly disagreeing over whether powerful AI models should be kept closed or made open for everyone to use. The divide is mainly between two groups. One group includes OpenAI and Anthropic, while the other includes Microsoft, Nvidia, Meta, IBM, Palantir and many AI start-ups.
Latest Update
The main reason behind this fight is China’s rapid progress in artificial intelligence. Chinese companies are quickly catching up with American AI firms through open-source development. Two Chinese AI start-ups—Z.ai and Moonshot AI—recently released AI models that reportedly rival those made by Anthropic and other American companies. These launches have increased concerns inside the US AI industry.
Also read: Americans switch to Chinese AI as cheaper models challenge ChatGPT and Claude despite Trump crackdown
China has embraced open-source AI more aggressively than the United States. Last year, DeepSeek released an open-source model that rivalled leading American systems, while companies such as Z.ai and Moonshot AI have continued to narrow the gap. Chinese President Xi Jinping has also promoted open-source AI as part of China’s global strategy.
People familiar with internal discussions said OpenAI and Anthropic fear they are losing their long-held lead. Both companies are lobbying Washington for tighter controls, arguing that AI models created through illicit distillation pose national security risks.
How China is catching up in AI
source
OpenAI and Anthropic believe that some advanced AI models are too dangerous to be freely available. They argue that these powerful models should remain under strict control because they could pose safety and security risks. Microsoft, Nvidia and many other tech companies believe AI should stay open-source. According to The New York Times, they say open-source AI allows more people to improve the technology, build new businesses and examine the models for security flaws.
Microsoft CEO Satya Nadella also backed open-source AI, calling it “essential to a healthy AI ecosystem.” He joined Jensen Huang and executives from Meta, IBM and Palantir in signing an industry letter supporting open-source AI, as noted by The New York Times.
The disagreement has now moved beyond Silicon Valley, with both sides taking their arguments to Washington. At the same time, OpenAI and Anthropic have been quietly lobbying the US government. Five people familiar with the discussions said the companies have warned Washington about Chinese open-source AI models. Treasury Secretary Scott Bessent and President Donald Trump’s science and technology adviser Michael Kratsios have both become involved in discussions about protecting American AI technology.
US officials are now examining possible action against Chinese companies. However, the US government has not decided on a blanket ban against Chinese AI models. Officials are instead considering dealing with individual Chinese models as national security cases. Supporters including Microsoft and Nvidia argue that open-source AI speeds innovation while allowing more researchers to identify security flaws.
Meanwhile, nearly 200 Silicon Valley start-ups under the Little Tech Association urged the Trump administration not to restrict Chinese open-source AI. More technology leaders have now publicly joined the open-source side.
Support for open-source AI has continued to grow across Silicon Valley. Elon Musk, Mark Zuckerberg and Google CEO Sundar Pichai joined Jensen Huang and Satya Nadella in backing open models, arguing they are essential for innovation and competition.
The fight over Chinese AI has turned a long-running debate about open versus closed AI into one of Silicon Valley’s biggest battles, with implications for innovation, national security and the future balance of power between the US and China.
