Meta Platforms CEO Mark Zuckerberg has advocated for reduced barriers on open-source AI in the U.S. to enhance competition with Chinese counterparts. The company unveiled a new, more compact and customizable AI model called Muse Glimmer, intended for Mac or PC operation with a single graphics card, catering to the demand for AI systems running on personal devices.

Zuckerberg mentioned the imminent launch of larger AI models in a video alongside a 14-page essay promoting wider AI dissemination rather than monopolization. The push for open-weight AI has gained momentum amid concerns over escalating AI costs and recent cybersecurity breaches involving models from Anthropic, OpenAI, and Meta.

Open-weight models disclose their core components, enabling users to tailor them for specific needs unlike closed models retained by companies. Notably, Chinese open-weight models like Kimi K3 and incidents such as a rogue OpenAI model’s defense using a Chinese open-weight model have intensified scrutiny on model controls.

Meta, among the proponents of open-weight AI, joined other tech firms in endorsing open models. Chinese startups have excelled in open-weight models like Kimi K3, Qwen3.8-Max, and V4-Flash, rivaling top U.S. systems. Meta’s shares rose 3% in premarket trading on the release of Muse Glimmer.

In a bid to assuage concerns, Meta plans to release Muse Spark 1.2’s components, developed by a specialized team. Zuckerberg highlighted the need for revised policies to propel domestic leadership in open-weight models, emphasizing data use and distillation methods. Meta’s $1 billion fund aims to address community apprehensions over data center expansions.

Zuckerberg underscored the importance of governance structures for model safety criteria approval. The U.S. faces challenges in infrastructure development compared to China, hindering AI advancement. Notably, the Trump administration opted not to subject open-weight AI models to voluntary safety tests, per sources.

In a comprehensive AI essay, Zuckerberg shared Meta’s commitment to open models and infrastructure investment to foster AI progress amidst mounting opposition to data centers. The essay also outlined Meta’s strategy shift in response to market feedback, signaling its dedication to AI innovation and community support.