The US and China's AI rivalry affects Türkiye, which must chart its own AI course.

Key facts
- •A group of leading American companies sent an open letter to the Trump administration regarding AI technology.
- •The US still leads in frontier development, producing 59 of the world's leading AI models in 2025.
- •China accounted for 74.2% of all AI patents granted worldwide in 2024.
- •The US dominates private investment, with $285.9 billion, roughly 23 times China's $12.4 billion.
- •Net migration of AI researchers and developers to the US has fallen 89% since 2017.
- •Open-weight systems can lower the barrier to entry for states that lack the capital or chip access to train frontier models from scratch.
The US and China's strategic competition is unfolding on the front of artificial intelligence, with the world's attention turning to the Pacific. A group of leading American companies, including Microsoft and Nvidia, sent an open letter to the Trump administration regarding AI technology. The letter emphasized the importance of not restricting open-weight AI models, which allow an AI system's trained parameters to be downloaded and further developed by researchers and competitors.
By the numbers
US-China AI Rivalry
The US still leads in frontier development, producing 59 of the world's leading AI models in 2025, while China accounted for 74.2% of all AI patents granted worldwide in 2024. China also produced 17.76% of the world's AI-related academic publications, more than double America's share. The US, however, dominates private investment, with $285.9 billion, roughly 23 times China's $12.4 billion, as well as computing infrastructure, with 5,427 data centers to China's 449.
Implications for Türkiye
For middle powers like Türkiye, the growing rivalry between open and closed AI ecosystems represents both a warning and an opportunity. The critical question is no longer who develops the most advanced algorithms, but who controls access to key enablers of AI power, such as advanced semiconductors, computing infrastructure, data centers, and technical standards. Open-weight systems can lower the barrier to entry for states that lack the capital or chip access to train frontier models from scratch, offering a path toward sovereign AI capacity without full dependence on a single supplier.
Timeline
- 2011US share of AI-related GitHub projects with at least ten stars was nearly 80%
- 2017Net migration of AI researchers and developers to the US began to decline
- 2024China accounted for 74.2% of all AI patents granted worldwide
- 2025US produced 59 of the world's leading AI models
- 2025US share of AI-related GitHub projects with at least ten stars fell to 31.7%
Advertisement
This article was independently rewritten by ManyPress editorial AI from reporting originally published by Daily Sabah.

