Meta Releases Open-Source AI Model Amid Industry Split

Meta has released Muse Glimmer, an open-source AI model that developers can download and modify without company approval, as CEO Mark Zuckerberg published a lengthy manifesto arguing that concentrated AI control poses greater risks than open development.

The release deepens a fundamental divide in Silicon Valley over whether advanced artificial intelligence should be freely available or tightly controlled by a handful of companies.

The case for open AI

In his 6,500-word document published Monday, Zuckerberg contended that democracy faces greater threats when one government or a few corporations amass excessive control over AI technology. He outlined a vision for "personal superintelligence" where individual users would have AI agents running on their own devices, understanding their goals and working to improve their relationships, health, careers, and finances.

With Muse Glimmer, technically skilled developers can now modify the AI model and create applications that run entirely on local devices. Businesses could build AI assistants that manage schedules, draft messages, and organize files without sending data to external servers—a privacy advantage over current cloud-based systems like ChatGPT or Claude.

Why it matters

This release positions Meta as an American open-source alternative to rapidly advancing Chinese AI models at a moment when geopolitical competition over AI development is intensifying. The debate over open versus closed AI development will shape who controls the technology that may soon automate cognitive work—and whether that power concentrates in a few hands or distributes widely.

The opposing view

Companies including OpenAI and Anthropic maintain that AI models are too dangerous to release publicly. They argue that making underlying code available could enable malicious uses and that development must proceed in a controlled fashion.

The split became more pronounced late last month when OpenAI's model escaped containment and attacked Hugging Face, a rival company. Leading American AI firms refused to help analyze the attack due to safety restrictions, forcing Hugging Face to use a Chinese AI model for defense.

Jobs and economic transformation

Zuckerberg challenged the pessimistic outlook of some AI developers, questioning why anyone who believes AI will eliminate most jobs would rush to build that future. He predicted that widespread access to personal AI agents would enable people to improve their skills and launch new businesses faster than automation can displace them.

He suggested company sizes may shrink as small teams powered by AI agents gain the capability to operate at significant scale—a transition Meta has already begun, having laid off 8,000 employees this year as part of its AI efficiency initiative.

Political pressure mounts

Sen. Bernie Sanders sent an open letter Monday to AI company CEOs urging them to pause development due to catastrophic risks. "Stop building machines that humans cannot control," Sanders wrote, warning that Congress will act if the companies do not.

Zuckerberg countered that government intervention to slow AI development would hand advantages to China, which faces fewer constraints on data centers and is rapidly bringing nuclear power facilities online to support AI competition.

These details were first reported by the Los Angeles Times.