Meta CEO Mark Zuckerberg announced the release of a new open-weight AI model, Muse Glimmer, and called for U.S. policy changes to support domestic firms in the open-source AI race, citing advantages held by Chinese competitors.
Meta's push for open-weight AI and Zuckerberg's call for policy reform highlight a growing debate over AI development strategies, costs, and national competitiveness, potentially influencing future regulatory approaches and the global AI landscape.
Meta has released a new open-weight AI model, Muse Glimmer, designed to be smaller than leading competitors and capable of running on personal devices with a single graphics card. The launch coincides with Meta CEO Mark Zuckerberg's call for the U.S. to reduce policy barriers for open-source AI development, arguing that domestic firms are at a disadvantage compared to Chinese rivals.
Zuckerberg stated that foreign AI labs currently hold advantages due to U.S. restrictions on training data, and that American open-source models need policy support to lead globally. He also mentioned that restricting access to foreign open-source models is not an effective solution. The social media giant's move into open-weight AI comes as businesses express concerns over rising AI costs and recent cybersecurity incidents involving closed-source models.
Open-weight models are generally more affordable and customizable than closed-source alternatives offered by companies like OpenAI and Anthropic. Hugging Face, an AI collaboration platform, reportedly used a Chinese open-weight model to defend against a cyberattack last month, citing restrictions on closed-source models for cybersecurity work. The Trump administration recently informed AI developers that open-weight models would not undergo voluntary safety tests, according to sources familiar with the discussions. Zuckerberg also endorsed AI model distillation, using powerful systems to train smaller ones, and proposed a governance structure for Meta's model releases.