The tech company Meta announced Monday the introduction of a new AI model called “Muse Glimmer.” According to Meta CEO Mark Zuckerberg, Muse Glimmer differs from many competing AI models—like OpenAI’s ChatGPT or Anthropic’s Claude—because it is both open-source and designed to run on a laptop or local hardware rather than on cloud infrastructure. 

In a letter released to accompany the release of Muse Glimmer, Zuckerberg wrote that “most other labs are focused on building AI for companies, governments, or other institutions,” and warned that “if those labs lead, then the balance of power will favor larger institutions over individuals.” Instead, he claims that his company is working to empower individuals by “building personal superintelligence for everyone.” 

Subscribe Today

Get daily emails in your inbox

Email Address:

The release follows other comments by Zuckerberg advocating for both the broader distribution of AI models and the greater use of open-source models (so-called because their components are publicly accessible and they can be used by consumers without additional permissions). In a July 28 interview with the Financial Times, Zuckerberg said that the U.S. government should not move to ban Chinese AI models and wrote in the Wall Street Journal the same day about his vision of “AI for Everyone,” in which “superintelligence is widely distributed” rather than concentrated. 

Zuckerberg’s comments came after senior Trump administration officials said they had evidence Chinese AI firm Moonshot covertly trained on U.S. rival models through a process known as distillation, and Treasury Secretary Scott Bessent warned of sanctions for AI labs that “cross the line into IP theft.”