Meta CEO Mark Zuckerberg is urging the US to loosen restrictions on open-source AI.
He argues that Washington risks falling behind China if American companies face too many barriers.
The call came as Meta unveiled Muse Glimmer, a smaller open-weight AI model.
It is designed to run agentic tasks directly on a Mac or PC using a single graphics card.
Zuckerberg hinted that bigger models are on the way, too.
Why the push for open AI? Cost and competition are big factors.
Open-weight models can be cheaper and easier for businesses to customise than closed systems controlled by companies such as OpenAI and Anthropic.
Zuckerberg argued that concentrating powerful AI in the hands of a few companies could create its own dangers.
“The notion AI is so dangerous that the only safe path is an extreme concentration of power seems inherently problematic,” he said.
Faster AI Developments
Chinese firms are making rapid progress in open-weight AI, increasing pressure on US developers.
Meta also plans to release the weights of Muse Spark 1.2, its most advanced model.
But AI’s growth comes with another headache: infrastructure.
Meta expects to spend up to $145 billion on AI infrastructure this year, while communities increasingly resist new data centres.
Zuckerberg says the US must make it easier to build both AI models and the infrastructure behind them.
The message is clear: if America wants to win the AI race, it may need fewer brakes and a lot more room to build.
