MENLO PARK, California: Meta Platforms unveiled a new open-weight artificial intelligence model on Monday as Chief Executive Mark Zuckerberg urged the United States to lower barriers for open-source AI, arguing that American developers need greater flexibility to compete with Chinese rivals.

The company released Muse Glimmer, a smaller AI model designed to perform agentic tasks on a Mac or PC using a single graphics card, targeting growing demand for AI systems that can run directly on users' devices.

"We've got even bigger models that are coming soon," Zuckerberg said in a video accompanying his 14-page essay, "The Future is for Everyone," in which he argued that AI should be widely available rather than concentrated in the hands of a few companies.

"The notion AI is so dangerous that the only safe path is an extreme concentration of power seems inherently problematic."

Meta's latest announcement adds to the growing support for open-weight AI models, which businesses increasingly view as a lower-cost alternative to proprietary systems, even as concerns grow over cybersecurity incidents involving models from Anthropic, OpenAI and Meta.

Unlike closed models from companies such as OpenAI, Anthropic and Alphabet's Google, open-weight models make their core components publicly available, allowing developers to customize them more easily.

The debate has gained prominence after Hugging Face, the AI coding collaboration platform that a rogue OpenAI model hacked, used a Chinese open-weight model to help defend against the attack because closed-source models impose restrictions on cybersecurity applications.

Chinese AI startups, including Moonshot, Alibaba, and DeepSeek, have emerged as leading developers of open-weight models, with Kimi K3, Qwen3.8-Max, and V4-Flash competing with top U.S. systems.

Meta shares, down about 10 percent this year, rose nearly three percent in premarket trading.

The company also said it plans to release the weights for Muse Spark 1.2, its most advanced model, developed by the superintelligence team it formed last year.

In his essay, Zuckerberg also announced a US$1 billion fund to support communities affected by the rapid expansion of AI data centers, which have faced growing local opposition across the United States.

"One significant disadvantage that the U.S. has compared to countries like China is that it is more difficult to build infrastructure here," Zuckerberg said. Meta expects to spend as much as $145 billion this year on AI infrastructure.

He also called for changes to U.S. policies on data use and AI distillation, saying they are needed if American companies are to remain competitive in open-weight AI.

"Foreign labs currently hold several advantages here since American labs have to comply with many additional restrictions on training data," Zuckerberg said.

"U.S. policy must reduce this additional friction if we want American open source models to lead over time," he added, arguing that restricting access to foreign open-source models was not an effective solution.

According to two sources familiar with the discussions, the Trump administration told AI developers earlier this month that it will not subject open-weight AI models to voluntary safety tests.