China's internet regulator unveiled a draft law that would require new generative AI products to undergo a security review before they're launched.
AI-created
content would also be required to "reflect core socialist values" and
"not contain content on subversion of state power," according to the
Cyberspace Administration of China (CAC) rules.
- Andy Chun, a City University of Hong Kong adjunct professor, told AFP that the draft law "is one of the strictest measures for generative AI so far."
- The
regulations would ban AI-generated content containing "terrorist or
extremist propaganda," "ethnic hatred," or "other content that may
disrupt economic and social order."
- Providers must ensure that
models don't output false information, and make sure training data for
AI doesn't discriminate on the basis of ethnicity, race, and gender,
according to the draft.
- Failure to comply could result in the suspension of services, fines, or criminal investigations, it said.
- Chinese tech giants Baidu, Alibaba, and SenseTime
all announced AI models that can generate text and images in recent
weeks, putting pressure on Beijing to rein in rogue generative AI
without stifling innovation in the emerging field.
