Meta Platforms CEO Mark Zuckerberg has urged for reduced barriers on open-source AI in the U.S. to enhance competition with Chinese counterparts. The social media giant unveiled a smaller, more customizable model called Muse Glimmer, tailored for running tasks on a Mac or PC with a single graphics card. Zuckerberg hinted at the forthcoming release of even larger models in a video post accompanying his 14-page essay advocating for broader access to AI.
The trend towards open-weight AI is gaining momentum as businesses seek cost-effective alternatives to closed models. These open-weight models provide users with the ability to customize core components, unlike closed models controlled entirely by companies. Recent incidents involving Chinese open-weight models like Kimi K3 and the use of such models by Hugging Face to defend against cyberattacks have highlighted the advantages of open-weight AI.
Notably, Meta, along with prominent tech firms, signed an open letter endorsing open-weight AI. While Chinese startups have excelled in developing open-weight models like Kimi K3 and Qwen3.8-Max, Meta aims to regain its competitive edge by releasing Muse Spark 1.2, its advanced model. Zuckerberg emphasized the need for infrastructure development to support AI advancements in the U.S., citing challenges compared to China.
To address concerns over data center expansion, Meta announced a $1 billion US fund to support communities affected by its operations. Zuckerberg stressed the importance of revisiting policies on data use and distillation to ensure U.S. firms lead in open-weight AI. Meta plans to implement a governance structure empowering independent directors to approve model releases.
In a bid to foster innovation and competitiveness in the AI landscape, Zuckerberg called for reducing regulatory hurdles for American open-source models. The Trump administration’s stance on open-weight AI safety testing was also highlighted, emphasizing the need for a conducive environment for American models to thrive.
