Taipei, Taiwan – After catching up with ChatGPT, China is racing to regulate the rapidly advancing field of artificial intelligence (AI).
A draft regulation released this week requires Chinese tech companies to register their generative AI products with China’s cyberspace agencies and submit them for security assessments before making them available to the public.
This regulation covers virtually all aspects of generative AI, from how it is trained to how it interacts with users. Elon Musk and Apple co-founder Steve Wozniak.
Under the rules announced by China’s Cyberspace Administration on Tuesday, technology companies will have to “validate the source of pre-training data” to ensure content reflects “core socialist values.” You will be responsible for your sexuality.
Businesses have used AI to call for the “overthrow of state power” and the overthrow of the ruling Chinese Communist Party (CCP), instigate moves to “divide the country” and “undermine national unity,” You should avoid creating content similar to Promote violence, extremism, terrorism, or discrimination.
It also restricts the use of personal data as part of generative AI training materials and requires users to verify their true identity before using the product.
Those who break the rules face fines ranging from 10,000 yuan ($1,454) to 100,000 yuan ($14,545) and possibly criminal investigation.
China has yet to match the success of California-based Open AI’s groundbreaking ChatGPT, but the move to regulate this nascent field is moving faster than elsewhere.
AI in the US is still largely unregulated outside of the recruitment industry. AI regulation has yet to receive much attention in the US Congress, but AI privacy-related regulations are set to roll out at the state level this year.
The European Union has proposed a comprehensive law, known as AI Law, that classifies which types of AI are “unacceptable” and prohibited, “high-risk” and regulated, and which are not regulated. bottom.
The law follows the EU’s General Data Protection Regulation 2018, which was passed in 2018 and is considered one of the toughest data privacy laws in the world.
Beyond the US and EU, Brazil is also working towards AI regulation, with a bill under consideration in the country’s Senate.
The proposed rule is still in the draft stage and open to public feedback until May, but following a broader regulatory crackdown on the tech industry that began in 2020, anti-competitive behavior and user data It covers everything, including the way it is. handled and stored.
Since then, Chinese regulators have introduced data privacy rules, created a registry of algorithms, and most recently started regulating deep synthesis technology, also known as “deep fakes.”
The regulatory push will ensure that “China’s big tech companies are going where the party state wants,” Chim Li, a China technology analyst at the Economist Intelligence Unit, told Al Jazeera.
Compared to other technologies, generative AI is a particularly thorny issue for the CCP, which is “concerned that these large-scale language models may generate politically sensitive content.” said Jeffrey Ding, an assistant professor at George Washington University who studies Chinese technology. Sector, told Al Jazeera.
Human-like chatbots like ChatGPT, which has been restricted in China, have been tapped by hundreds from across the internet, including topics Beijing considers taboo, such as Taiwan’s disputed political status and the 1989 crackdown on Tiananmen Square. Discard 10,000 data points.
In 2017, two early Chinese chatbots went offline after telling users they didn’t like the Chinese Communist Party and wanted to move to the United States.
ChatGPT, released in November, has also sparked controversy in the West, from telling users who pretended to be mentally ill to commit suicide to encouraging a New York Times journalist to break up with his wife.
The answers to questions created by ChatGPT impressed many users, but they also contained problems such as inaccurate information and broken URL links.
ChatGPT’s Chinese competitors, like Baidu’s ERNIE, train using data from outside China’s “Great Firewall,” including information gleaned from banned websites such as Wikipedia and Reddit. I was. ERNIE has been widely viewed as inferior to ChatGPT, despite its access to information Beijing considers sensitive.
Beijing’s rules on AI could create major implementation headaches for companies like Baidu and Alibaba, said ChatGPT rival Tongyi Qianwen this week, said Matt, a researcher at the Carnegie Endowment for International Peace. Sheehan told Al Jazeera.
Sheehan said the regulations set “very high standards” and it was unclear whether companies would be able to meet them with the technology currently available.
Regulators may choose not to enforce the rules strictly at first unless they find particularly egregious violations or decide to set an example for a particular company, Sheehan added.
“Like many Chinese regulations, they define things pretty broadly, so they essentially shift power to regulators and enforcers, allowing them to enforce and punish companies when they choose. This is especially likely, he added, if they produce “inaccurate” output that contradicts official government explanations.
