Ant Group warns that US tech giants are using open-source tools to bind developers to closed AI ecosystems, while Chinese firms are opening their models to drive broader adoption.
Ant Group has spotlighted a deepening divide in the global AI landscape, releasing its report on open source trends at the Inclusion Conference on the Bund in Shanghai. The report claims that American tech giants, including OpenAI and Nvidia, are attempting to “lock in developers to their closed-source artificial intelligence ecosystems by open-sourcing tools in other layers of the tech stack.”
Most leading US large language models remain closed-source. In contrast, Chinese players such as Alibaba Cloud and TikTok-owner ByteDance have chosen to open-source their models, enabling developers to download and build on them directly. This difference underscores two opposing strategies in the open-source battleground: while US firms release selective tools, Chinese firms release full models.
A tech stack in AI spans both hardware, such as semiconductor chips, and software, such as algorithms and frameworks. According to Ant Group, US companies have focused their open-source efforts on AI development “toolchains” to encourage adoption of their proprietary models and hardware.
One example is Nvidia’s Dynamo, open-sourced in March. The inference platform is optimised for deploying large-scale AI models and marketed as the ‘operating system of AI.’ While Dynamo integrates with open-source frameworks such as PyTorch and SGLang, it is primarily designed to pair with Nvidia’s powerful GPUs, Ant Group noted.
The report highlights a strategic divergence: US companies selectively open-sourcing tools to strengthen reliance on proprietary ecosystems, and Chinese firms openly releasing models to foster community-driven innovation.
