Meta’s CEO, Mark Zuckerberg, has appealed to the U.S. government to loosen regulations surrounding open-source AI models. He contends that AI technology should be made broadly available, rather than being confined to the control of a select few entities. His comments have ignited a lively debate within the open-source AI community.

Meta’s AI model, Muse Glimmer, is designed to operate locally on Macs or PCs equipped with 24GB or 32GB of VRAM. Amid escalating AI development costs and recurring cybersecurity issues at closed-source AI firms, open-weight AI models are drawing increasing interest due to their cost-effective training and adaptability. In the realm of open-weight models, Chinese companies have demonstrated remarkable prowess. Notably, Moonshot AI’s Kimi K3, Alibaba’s Qwen3.8-Max, and DeepSeek’s V4-Flash are all capable of rivaling leading U.S. AI models.