But since then, predictions about Apple’s level of capitulation have only gotten harder. Algorithms that decide what the public sees online or through AI will have to be registered with Chinese authorities, and the new AI law will largely focus on the exact public interfaces that Western tech companies want to engage with. It focuses on controlling the model.
“You have to submit to regulators. You may have to submit a lot of details about coding and so on…many technology companies may not want to do that,” Tan says.
The problem is that China can afford to take such steps because the balance of power is now more in its favor than ever.
“China is no longer just playing a follower role in many technology fields; it has already advanced and is playing a leading role,” Tan added.
Is it business as usual?
From a Western perspective, the rules set for China’s generative AI oscillate between the laudable and the alarming.
“This regulation includes many ambiguous censorship requirements, such as that deep-synthesized content must “follow the correct political direction,” must not “disturb economic or social order,” and must not be used to generate fake news. ” states the Carnegie Endowment paper. Situation in 2023.
“Deep synthesis” is the term CAC uses instead of generative AI. Due to Chinese restrictions, Siri will not talk about the Dalai Lama, refer to Taiwan as a separate country, or acknowledge the Uyghur people. Who knows what else is out there.
Given the current lax state of Western LLMs, it’s pretty hard to imagine a chatbot that can’t be led to believe that China is part of the sovereign nation of Taiwan, much less agree with it 100 percent of the time. But it is clear that many Chinese tech companies have managed to comply, at least to the satisfaction of regulators. In August 2024, the South China Morning Post reported that 188 LLMs had been approved for use so far, compared to just 14 in January 2024.
You could also argue that it’s business as usual for Apple to effectively employ a custom version of one of these LLMs to fill out the Chinese version of Apple Intelligence. Apple already censors its app store to comply with Chinese policies. We are already working with local organizations.
But with Apple Intelligence-generated AI at the heart of iPhones and other devices, the company has been accused of being too embedded as an American company seeking reassurance at the wishes and whims of the Chinese state. It appears that the risk of
In August, Zhuang Rongwen, director-general of China’s Cyberspace Administration, said generative AI such as chatbots is “powerfully driving economic and social growth.” A 2021 report in the New York Times suggested that China’s iPhone user data is not actually needed to monitor its citizens because the government already has more powerful methods. However, with the introduction of GenAI, Apple may inadvertently become a more active participant in CCP’s goals.
