Huang’s reported exit clouds Nvidia’s open-model push
Jensen Huang’s reported departure from Nvidia leaves uncertainty over an open-weight AI strategy designed to reduce businesses’ dependence on proprietary providers such as Anthropic and OpenAI. Through its Nemotron models, Nvidia has been developing systems that companies can download, customize and run on their own infrastructure. That approach could challenge recurring access fees while increasing demand for Nvidia’s computing hardware.
The company launched its Nemotron Coalition with participants including Mistral AI, Perplexity, Cursor, LangChain, Reflection AI and Thinking Machines Lab. Nvidia was also reportedly pursuing investments in independent developers and an acquisition of Hugging Face. These efforts targeted enterprise applications, autonomous agents and specialized reasoning. Anthropic, meanwhile, has been developing software compatible with chips from AMD, Google and Amazon, potentially weakening Nvidia’s hardware advantage.
Huang reportedly resigned after personal misconduct, which he acknowledged while expressing willingness to return. His departure comes as Anthropic prepares for a potential public listing valued at up to $1.4 trillion. Anthropic has committed $40 million to Public First Action, which advocates stronger AI safeguards and oversight. There is no evidence Anthropic played a role in Huang’s removal, and Nvidia could continue developing its open models under new leadership.