The U.S. and China are presently at the forefront of artificial intelligence dominance, with the competition for top position-currently occupied by the U.S.-apparently just beginning to heat up.
As an enthusiast, I find myself marveling at the centerstage in this AI revolution across the USA, companies such as OpenAI (the geniuses behind ChatGPT), Meta, Alphabet (Google’s parent company), Microsoft, and NVIDIA have been trailblazers, with NVIDIA even breaking ground by surpassing a staggering $4 trillion valuation. On the other side of the globe in China, it’s companies like Baidu, Alibaba, Tencent, and ByteDance that are making significant strides to close the gap, eagerly joining the global AI race.
During an interview with CNBC on August 18, Sam Altman, CEO of OpenAI, expressed concern that the advancements in AI by China may be underestimated by the United States. Moreover, he suggested that export controls might not be effective in limiting China’s progress in this field, as the competition goes beyond just who is currently leading in the race.
These remarks follow a series of news stories about restrictions on the sale of AI technology equipment, particularly the highly desired NVIDIA H20 AI GPUs, which have garnered interest from Chinese businesses.
Regarding advancement, it’s not just about who is in the lead, but rather considering the intricacies. China potentially has the ability to progress faster due to its inference capacity and extensive research, which encompasses multiple layers from the initial stages of research to the development of products. It’s a complex scenario that can’t be easily summed up as a simple comparison between the U.S. and China.
OpenAI CEO Sam Altman, speaking to CNBC
As an analyst, I’d rephrase that event as follows: In April 2025, the US government enacted a policy prohibiting the export of NVIDIA H20 chips. However, this ban was lifted in July, following a visit by NVIDIA CEO Jensen Huang to the White House. It’s been reported that during this meeting, a deal was struck with President Trump, allowing for new export licenses to be issued for the AI hardware in question.
Following this, NVIDIA announced they would require an additional 300,000 H20 AI GPUs from TSMC to fulfill demand in China. This requirement was on top of the existing 600,000 to 700,000 chips that were already stored and waiting for a purchaser.
On August 10, it was announced that NVIDIA and AMD had agreed to hand over 15% of their earnings from particular AI chip sales – the NVIDIA H20 and the AMD MI308 – to the Chinese government, as a condition for obtaining the necessary export licenses for these sales.
Meanwhile, events unfolded, there were rising apprehensions about the export of security-related technology from both nations. In the United States, authorities aim to avoid providing China with AI chips, fearing they might strengthen their military or give them an edge in the global AI competition.
On China’s part, there’s a growing concern among officials regarding potential backdoors or hidden spyware within the hardware. This issue prompted NVIDIA’s Head of Security to pen an article entitled “Absolutely No Backdoors. No Kill Switches. No Spyware.
Sam Altman isn’t the first CEO of a major AI firm to express doubts about export control
A key point made by NVIDIA CEO Jensen Huang when advocating for the sale of H20 AI GPUs to China was that restrictive export policies may not be the solution in the race for artificial intelligence advancement.
He pointed out that it could be a significant error for non-American AI chips to establish themselves as the worldwide norm, leading eventually to China’s challenging market becoming dependent on local companies such as Huawei.
Sam Altman, in his interview with CNBC, expresses a similar understanding.
It seems you can regulate one aspect, but perhaps not the desired one; people might construct factories elsewhere or discover alternative methods. Frankly, it sounds challenging to find an straightforward solution. However, I have a feeling it’s going to be difficult.
OpenAI CEO Sam Altman, speaking to CNBC
Altman might be alluding to the thriving underground trade of AI chips, when he uses the term “alternative methods” or “unofficial channels”.
Based on information from the Financial Times, there are estimates suggesting that Chinese firms may have illicitly imported around a billion dollars’ worth of NVIDIA’s AI chips into the country within the initial three months after President Trump imposed an outright export prohibition on the H20 GPU in April.
Expanding on their previous criticism of NVIDIA, Gamers Nexus – known for its candidness towards NVIDIA’s business practices – recently released an extensive 3.5-hour undercover investigation focusing on the illegal GPU trade in China.
China’s officials are urging AI companies to avoid using NVIDIA’s H20 GPUs due to national security concerns, and it is becoming increasingly inappropriate for Chinese companies to continue using foreign hardware in this sector. However, an outright ban on NVIDIA’s products has not yet been imposed.
As a analyst, I can express this in a more personal and easy-to-understand manner: While our domestic options may not match NVIDIA’s hardware prowess at the moment, they have the potential to catch up. Alternatively, Huawei’s AI clusters could be a game-changer, potentially propelling China towards supremacy in this field, as suggested by Tom’s Hardware.
China is influencing OpenAI’s release schedule
On August 7, OpenAI unleashed its new GPT-5 model, but it was the subsequent criticism that drew significant attention. In the days leading up to the GPT-5 debut, OpenAI had actually rolled out two open-source AI models: gpt-oss-120b and gpt-oss-20b.
Since 2019, this is the first time the company has released open-source solutions. These tools are engineered to provide advanced and efficient chain-of-thought reasoning, while also offering high levels of customization.
What influenced OpenAI’s decision to release these open models? Sam Altman explains:
It became evident that, by not participating, the globe would predominantly adopt Chinese open-source frameworks. This consideration undeniably played a significant role in our decision-making process, though it wasn’t the sole factor. However, its impact was substantial.
OpenAI CEO Sam Altman, speaking to CNBC
Read More
- Minecraft lets you get the Lava Chicken song in-game — but it’s absurdly rare
- Gold Rate Forecast
- PS5’s ChinaJoy Booth Needs to Be Seen to Be Believed
- Wrestler Marcus “Buff” Bagwell Undergoes Leg Amputation
- Lewis Capaldi Details “Mental Episode” That Led to Him “Convulsing”
- Cyberpunk 2077’s Patch 2.3 is Here and It’s Another Excellent Overhaul
- Microsoft is on track to become the second $4 trillion company by market cap, following NVIDIA — and mass layoffs
- Elden Ring Nightreign’s Patch 1.02 update next week is adding a feature we’ve all been waiting for since launch — and another I’ve been begging for, too
- AI-powered malware eludes Microsoft Defender’s security checks 8% of the time — with just 3 months of training and “reinforcement learning” for around $1,600
- Rob Schneider’s Happy Gilmore 2 Role Is Much Different Than We Thought It’d Be
2025-08-20 14:10