Anthropic CEO Dario Amodei and Nvidia CEO Jensen Huang are warning the United States government against implementing blanket bans or broad crackdowns on open-weight artificial intelligence systems. This position comes as the government debates new AI rules and regulations.
In a recent blog post, Amodei categorically denied that Anthropic had ever advocated for a ban on open-weight models, breaking a silence that had led to accusations that the company wanted such a ban to protect its business. Amodei stated that open-weights models that do not have dangerous capabilities are a public good.
The warning from these Silicon Valley leaders follows a breakthrough Chinese AI model that has challenged US leadership in the technology. Rather than a general crackdown, Anthropic and Nvidia are urging regulators to instead focus on specific risks and misuse.
An open-weight model is defined as one where the developer releases its trained parameters, known as the weights, which are the numerical components of the system.