Meta and Nvidia Plant 'Very Firm Flag' in Open-Weight AI Race Led by Chinese Labs Last month, American tech giants urged policymakers to avoid premature restrictions on open-weight AI models, even those from China. Now, two of those companies are intensifying their efforts to compete by launching their own open-source offerings. The move comes amid growing tensions over the role of Chinese AI models and the practice of distillation, which critics argue could infringe on intellectual property rights. Industry leaders argue that restricting open-source models would harm innovation and concentrate power among a few firms. In an open letter dated July 24, a consortium of tech companies emphasized that open-weight AI could drive economic growth, strengthen competition, and maintain American technological leadership. They described distillation—a technique used to refine models—as a standard practice for improving and validating AI systems. This stance reflects broader industry concerns about balancing national security risks with the benefits of open collaboration. Meta and Nvidia have now taken concrete steps to advance their positions. On Monday, Meta unveiled Muse Glimmer, part of a strategy to release its most advanced AI models to the open-source community. CEO Mark Zuckerberg announced that the company would open the weights for its latest AI model, Muse Spark 1.2. Weights refer to the mathematical rules and calculations that govern how an AI operates. This marks a significant shift for Meta, which previously faced criticism for its Llama series, particularly after the release of Llama 4 in April 2025, which disappointed developers. A day later, Nvidia introduced Nemotron 3.5 Lightning, a model derived from its Nemotron 3 family, which was launched in December.#nvidia #meta #mark_zuckerberg #umesh_sachdev #charlie_dai
