HONG KONG, July 13 (Reuters) – China on Thursday announced measures to control its burgeoning generative artificial intelligence (AI) industry, softening the tone from previous drafts and allowing regulators to help develop the technology. said to aim for
The rules are set to go into effect on Aug. 15, what the Chinese government calls “temporary,” after officials signaled an end to a long-standing crackdown on the tech industry that has sought to fuel a slower-than-expected economic recovery. was done. After COVID-19 control measures have been lifted.
Analysts say the measures are far less burdensome than those outlined in the April draft, and the final rule says China wants to support the technology while ensuring its security. He said that consideration was also given to highlighting
A statement Thursday by the China Cyberspace Administration (CAC) said only providers wishing to offer services to the general public would be required to submit security assessments, giving room to companies developing products for enterprise use. suggested that
“The draft regulations had untenable standards regarding training data quality and output accuracy, essentially calling for perfection. ‘measures’,” said technology analyst Tom Nullist. At research firm Trivium China.
Investments are pouring into China’s generative AI scene, with companies like Baidu (9988.HK) and Alibaba Group (9988.HK) releasing dozens of AI models. But companies have held back from releasing chatbots to the public until the Chinese government finalizes regulations on the technology and approves products.
Instead, most Chinese tech companies have so far focused on finding industrial applications. On Thursday, JD.com was the latest to enter the race, announcing a large-scale enterprise language model called ChatRhino.
“The current version is very much in line with market expectations,” said Morningstar analyst Kai Wang.
“This sends a positive signal that the regulator is paving the way for Chinese companies to launch their products at scale.”
“socialist values”
China sees AI as an area it wants to rival the United States and aims to become a world leader by 2030.
The technology is seen to be ahead of regulation as countries around the world grapple with setting guardrails for the technology popularized by the success of OpenAI’s ChatGPT chatbot.
Such efforts must weigh security concerns and copyright protection against ensuring a conducive environment for innovation.
The CAC’s statement Thursday reiterated that content generated by generative AI for the public should be in line with China’s core socialist values.
In addition, service providers should ensure that their intellectual property rights are not infringed and should use legitimate data sources, he advised.
Citing areas such as generative AI algorithms and semiconductors, he said China would like to promote the development of this technology as well as work on the formulation of international rules.
“Relevant national authorities must improve their supervisory approaches to be scientific and compatible with innovation and development,” the regulator said.
Reported by Josh Ye.Editing: Brenda Goh, Clarence Fernandez, Angus McSwan
Our standards: Thomson Reuters Trust Principles.
