China Mandates Security Reviews for AI Services Like ChatGPT (bloomberg.com) 11
China plans to require a security review of generative AI services before they're allowed to operate, casting uncertainty over ChatGPT-like bots unveiled by the country's largest tech companies including Baidu. From a report: Providers of services must ensure content is accurate and respects intellectual property, and neither discriminates nor endangers security, the Cyberspace Administration of China said in draft guidelines published for public feedback. AI operators must also clearly label AI-generated content, the country's internet overseer said in a statement posted on its website.
The CAC's requirements add to Beijing's growing attempts to regulate the explosive growth of generative AI since OpenAI's ChatGPT fired up the industry in November. Companies from Alibaba Group to SenseTime and Baidu all aim to build the definitive next-generation AI platform for the world's largest internet market. That mirrors a growing wave of development abroad with Alphabet's Google and Microsoft among the many tech companies exploring generative AI, which can create original content from poetry to art just with simple user prompts. China's made no secret of its wish to elevate AI at a time the country is locked in a conflict with the US over technology from chips to EVs. But it remains uncertain how the government intends to both galvanize and police the emergent field.
The CAC's requirements add to Beijing's growing attempts to regulate the explosive growth of generative AI since OpenAI's ChatGPT fired up the industry in November. Companies from Alibaba Group to SenseTime and Baidu all aim to build the definitive next-generation AI platform for the world's largest internet market. That mirrors a growing wave of development abroad with Alphabet's Google and Microsoft among the many tech companies exploring generative AI, which can create original content from poetry to art just with simple user prompts. China's made no secret of its wish to elevate AI at a time the country is locked in a conflict with the US over technology from chips to EVs. But it remains uncertain how the government intends to both galvanize and police the emergent field.
need source code and the EULA says that we copy (Score:3)
need source code and the EULA says that we can copy it and you can only sue us in an china court.
Re:need source code and the EULA says that we copy (Score:4, Insightful)
No, it's not about the source code or anything.
They got to make sure it's not going to suggest things like "Freedom" or "democracy" or human rights. Or pesky things like a certain massacre, or genocide of certain muslims.
Those security problems.
Re: (Score:1)
all the data (Score:2)
all the data, anytime we want, no questions asked - CCP
Speecial rule (Score:1)
To avoid dangerous and subversive content the AI training data must not include the works of A.A.Milne.
Accuracy? (Score:4, Interesting)
Would love to ask it how many people were killed by the CCP in Tiananmen Square.
How would such a review be done? (Score:2)
is accurate ... nor endangers security
China is a sovereign nation and can do whatever is wants. The really interesting aspect to this story is how they intend to conduct the software review. If they can find a way to ascertain accuracy and characteristics of desired answers, that would be a huge software engineering breakthrough. It's already very challenging for traditional complex software systems, but AI systems are much more opaque. Source code review only looks at code structure and not the important model parameters. Practical valida
Re: (Score:2)
Sensible precaution (Score:2)
If it was not "greed above everything" in the west, we would have that as well.
Re: (Score:2)
You want to regulate Internet chat bots?