Meta has introduced a new open-weight AI model as CEO Mark Zuckerberg renewed his call for fewer US restrictions on open-source AI. The move comes as the company looks to strengthen its position in the global AI race and compete more effectively with Chinese developers.
The new model, Muse Glimmer, is smaller than leading models from competitors and is designed to run agentic tasks on a Mac or PC using a single graphics card. Meta said more and larger models are expected soon.
“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”. He argued that AI should be widely accessible rather than controlled by a small number of organisations.
“The notion AI is so dangerous that the only safe path is an extreme concentration of power seems inherently problematic.”
Open-weight AI models are gaining attention as businesses look for lower AI costs and greater flexibility. Unlike closed models, they provide publicly accessible core components that can be customised. Recent cybersecurity incidents involving AI models from Anthropic, OpenAI and Meta have also increased interest in alternative approaches.
Chinese companies are currently making strong progress in open-weight AI, with Moonshot’s Kimi K3, Alibaba’s Qwen3.8-Max and DeepSeek’s V4-Flash competing with leading US systems. OpenAI, Anthropic and Google’s leading models remain closed-weight.
Meta shares rose nearly 3% in premarket trading on Monday, despite falling about 10% so far this year. The company also plans to release the weights of Muse Spark 1.2, its most advanced model, developed by its superintelligence team.
Meta has supported open models for years, but the poor reception of Llama 4 last year had led the company to reconsider its strategy.
Zuckerberg also announced a new $1 billion fund to support communities affected by the company’s data-center expansion. Opposition to data centers has become a major US political issue and a challenge for Big Tech’s AI infrastructure plans.
Meta expects to spend as much as $145 billion on AI infrastructure this year. Zuckerberg said the US needs to reconsider policies around data use and distillation to help American open-weight models compete globally.
Meta will also establish a governance structure allowing independent directors to approve safety criteria for releasing models.
“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, while arguing that restricting foreign open-source models would not be effective.
The Trump administration earlier told AI developers that it would not subject open-weight AI models to voluntary safety tests, according to 2 sources familiar with the discussions.
Also read: Viksit Workforce for a Viksit Bharat
Do Follow: The Mainstream LinkedIn | The Mainstream Facebook | The Mainstream Youtube | The Mainstream Twitter
About us:
The Mainstream is a premier platform delivering the latest updates and informed perspectives across the technology business and cyber landscape. Built on research-driven, thought leadership and original intellectual property, The Mainstream also curates summits & conferences that convene decision makers to explore how technology reshapes industries and leadership. With a growing presence in India and globally across the Middle East, Africa, ASEAN, the USA, the UK and Australia, The Mainstream carries a vision to bring the latest happenings and insights to 8.2 billion people and to place technology at the centre of conversation for leaders navigating the future.