US pushes AI safety, China backs open source, where

What happened

History appears to be repeating itself, with the world once again divided into opposing camps. The latest wave of debate began with the resignation of Anthropic researcher Jacob Coxon, who warned in a post on X that leading AI companies were advancing the technology irresponsibly and that increasingly powerful AI systems could pose a lethal threat to humanity. “I resigned from Anthropic today.

There is no doubt that the US and China are two major pillars of global technological innovation and advancement. But when the world’s two technological powers take sharply divergent approaches to a frontier technology like artificial intelligence, a larger question emerges: could this divide hamper the global growth of AI, particularly as its development demands enormous resources while its economic, technological and societal consequences extend far beyond national borders?

“US debate increasingly focuses on whether safety mechanisms can keep pace with frontier AI development amid intense commercial and geopolitical competition, China is emphasising open-source development and wider adoption to expand its technological influence. At the same time, Beijing has also called for stronger regulation, monitoring and safeguards to keep AI under human control,” said Kumar.

Why it matters

I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly,” Coxon wrote.

After Coxon, Joe Benton, who used to lead a safety research team at Anthropic, and Josh Engels, who used to work on AI safety research at Google also left their companies citing AI’s impact of controlling human lives which resulted in the big AI companies CEO ‘s talking about pacing means slowing down the AI development to make sure AI will not get out of the hand of human control.

While one side of the AI industry was calling for a slowdown in development and greater focus on safety, a very different vision was taking shape at the BRICS 2026 summit hosted by India.

Chinese President Xi Jinping proposed building a BRICS open-source AI ecosystem, and promoting deepened cooperation and push for shared resources among member countries in developing and deploying large language models, AI training and a common digital cloud platform. Several Indian startups are building homegrown foundational models, data centre capacity is expanding, and the country is developing its own datasets and language AI infrastructure through initiatives such as Bhashini.

What to watch

“From a technical perspective, this means that we could potentially, on the fly, generate models given the application. The current debate over AI safety is not entirely hypothetical. As Nittur pointed out, AI’s capabilities are immense, making guardrails essential to ensure that innovation continues while the technology remains under human control. The debate is also increasingly viewed through a geopolitical lens.

“The competition may not simply be about who develops the most powerful model, but whose AI ecosystem the rest of the world adopts. China is betting heavily on diffusion and on building an ecosystem that can spread internationally. How Does India Navigate the US-China AI Divide? India’s position is particularly significant in this global AI debate, given the sheer scale of its AI user base. By total usage volume, India is the world’s second-largest AI market, behind only the United States.