Cathie Wood Says Open Source AI Is Making OpenAI and Anthropic Richer, Not Poorer

The common view says open-weight AI is bad news for frontier labs like OpenAI and Anthropic. ARK Invest's Cathie Wood argues the opposite. An open-weight AI model is one in which the creator releases the underlying parameters publicly. It allows anyone to download, run, and modify it on their own hardware without paying for access.

Wood says the real threat of open-weight models is not that they replace frontier AI.It is that bad actors use them to attack enterprises. That security pressure, she argues, forces companies to keep buying frontier-grade AI for protection. This makes Open AI and Anthropic richer, not poorer.

Some Models Are Arming the Threat

Wood's case inverts the popular investment narrative. As open-weight models from Meta, Mistral, and DeepSeek grow more capable, the conventional view is that enterprises will shift away from expensive frontier subscriptions.