Skip to content

China Decode: The U.S. vs China AI Battle Is Getting Ugly (Transcript)

Editor’s Notes: In this episode of China Decode, Alice Han and James Kynge examine the escalating AI rivalry between the U.S. and China, highlighted by White House accusations of industrial-scale intellectual property theft. The discussion covers significant market shifts, including U.S. banks increasingly borrowing in Chinese currency through the “dim sum” bond market. They also explore the curious social phenomenon of “pretend to work” offices in China, which reflects a deepening youth unemployment crisis and changing cultural attitudes toward work. The episode concludes with bold predictions regarding the future of offshore Chinese debt and potential new inheritance tax laws as the country prepares for a massive intergenerational wealth transfer. (April 28, 2026)

TRANSCRIPT:

Introduction

JAMES KYNGE: Now we’ve got the White House accusing China of industrial-scale theft of intellectual property from US AI labs. This is different. And on the China side, we’ve got the Chinese government banning the acquisition by one of America’s biggest companies, Meta, of a promising AI company that was founded in China but is now based in Singapore. So, I would say all of this to me means that the rivalry has entered a new phase.

ALICE HAN: Welcome to China Decode. I’m Alice Han.

JAMES KYNGE: And I’m James Kynge.

Today’s Topics and Market Update

ALICE HAN: In today’s episode of China Decode, we’re discussing the tightening and increasingly combative race for global AI supremacy, why US banks are scooping up Chinese currency in offshore markets, and the growing industry of pretend-to-work offices in China.

That’s all coming up, but first, let’s do a quick check-in with how the Chinese markets are starting the week. On Monday, the markets opened the week with slight gains, the Shanghai Composite closing up 0.16%, while the Shenzhen Component was up 0.37%. Industrial profits jumped 15.8% in March year-on-year, with enterprise profits growing by over 15% in Q1. That growth comes in spite of the shocks brought on by the war in Iran and is largely attributed to the booming AI and chip industries in mainland China. The high-tech manufacturing sector alone saw a 47.4% gain in profits in Q1.

Stock market standouts included More Threads, up 8.5%, and SMIC A-shares, up 5%. Fibercom Wireless and Liad Optoelectronic Company were both down over 12%. And tech giant stocks were broadly mixed, with Tencent ending the day down 3% and Baidu up 3.5%.

The AI Race Escalates: Meta, Manus, and Industrial-Scale Theft

ALICE HAN: All right, let’s get into it. As the US and China barrel towards a high-stakes summit next month, the fight over artificial intelligence is getting a lot more aggressive and a lot more complicated. Just today, China has announced it will prohibit foreign investment in Manus, a Singapore-based AI company with Chinese founders, which means Meta’s December acquisition of the company will have to be undone. Experts are saying the move could hold other Chinese entrepreneurs, especially in the AI space, from seeking out foreign partners.

Meanwhile, the White House has now accused China-backed actors of running industrial-scale campaigns to essentially siphon off the capabilities of American AI labs, querying systems millions of times to replicate how they work. It’s the kind of claim that is hard to definitively prove, but if true, it suggests the global AI race isn’t just about innovation, it’s about extraction.

And here’s where the timing gets really interesting. Just as those accusations ramp up, Chinese startup Deepseek has rolled out a powerful new model built cheaper, released to open source, and quickly closing the gap with the US AI leaders. What some are calling China’s Deepseek moment is starting to look less like a one-off and more like a real strategy.

James, we talk a lot about AI on this podcast, and with good reason, I think. I think some of the news that has come out seems to suggest that there is an escalation now in this AI conflict between the US and China, and it’s not just driven by Washington and Beijing. I detected in China — I was just there for 3 weeks — consternation about AI competition, but also about the politics of Silicon Valley. I want to get your take really quickly on whether or not AI is going to be perceived broadly as a threat, not just by American legislators, but also by Silicon Valley.

A Cold War Curtain Over AI Competition

JAMES KYNGE: Well, welcome back, Alice. Welcome back from China. I’m dying to hear more about your trip there. But as far as I see it, this really is pretty big. We’ve got what I would say is kind of like a Cold War curtain coming down over the US-China competition over AI.

Until now, there’s been a lot of rivalry, mostly commercial rivalry. We had the OpenAI ChatGPT in late 2022, and then we had the first Deepseek moment in early 2025. And both the US and China were going really strongly, fueled by a lot of money and a lot of enterprise. But now we’ve got the White House accusing China of industrial-scale theft of intellectual property from US AI labs. This is different. And on the China side, we’ve got the Chinese government banning the acquisition by one of America’s biggest companies, Meta, of a promising AI company that was founded in China but is now based in Singapore. So I would say all of this to me means that the rivalry has entered a new phase.

The White House is talking about exploring measures to hold foreign actors accountable — that’s the phrase they use — for “industrial-scale distillation campaigns.” To me, I think the important thing is that politically, geopolitically, this is a step change. It is like the US law enforcement agencies are limbering up to really get moving on the AI frontiers with China.

So what is distilling? It’s basically the scenario in which a Chinese AI company would send hundreds of thousands or millions of prompts, probably from proxy accounts through VPNs to camouflage their source, to a US LLM — that’s a large language model.