China will approve the first batch of generative artificial intelligence services for public rollout as soon as this week, freeing up homegrown technology champions including Baidu Inc. to openly compete with the likes of OpenAI Inc. and Microsoft Corp.
The Cyberspace Administration of China will soon give the greenlight to a handful of firms including fledgling players and major technology names, people familiar with the matter said. Search leader Baidu is expected to be among the first to win approval from Beijing’s top internet watchdog, according to one of the people, who asked not to be identified because the information isn’t public.
The approvals, which are coming just weeks after the government implemented sweeping new regulations governing AI, will officially bring ChatGPT-style services to one billion or more internet users for the first time. Baidu was up 5.5% in New York on Wednesday.
The move marks a milestone for China’s budding AI industry and its would-be leaders, which since the debut of OpenAI’s ChatGPT have plowed billions of dollars into developing services capable of competing with the likes of Microsoft Corp. and Alphabet Inc.’s Google.
Beijing deems AI both a business and political imperative given the transformative nature of the technology. Baidu has so far stolen much of the limelight by debuting Ernie Bot in March, but Alibaba Group Holding Ltd. and Tencent Holdings Ltd. have since joined a parade of startups in touting their own foundation models.
It’s not immediately clear which companies other than Baidu were among the first to win government approval to be providers of AI models, the people said, asking not to be identified discussing private information. The CAC didn’t respond to a faxed request for comment. A spokesperson with Baidu didn’t respond to an emailed request for comment.
Beijing has designated AI one of a dozen tech priorities and, after a two-year regulatory crackdown, the government is seeking private sector help to prop up a flagging economy. China, like Europe, quickly moved ahead with regulatory oversight of what may be the most promising — and controversial — technology of the last 30 years, while the US has no legislation under serious consideration even after industry leaders warned that AI poses a “risk of extinction.”
The Chinese rules that kicked in Aug. 15 required generative AI services to get clearance before they become available to domestic users, part of guardrails intended to control content while allowing players to flourish and compete with US rivals.
Services with content that can potentially influence public perception are required to pass a security review and file their algorithms for record-keeping, the rules said. But such guidelines were part of a watered-down, finalized version of the regulations first unfurled in April, signaling Beijing’s endorsement of the industry.
In a post-earnings call in August, Baidu’s billionaire co-founder Robin Li said the company was working closely with regulators to push the proper use of generative AI, while skirting questions about an exact launch date for Ernie.
It’s still too early to tell whether Chinese-made generative AI services could rise to the level of must-have apps like Tencent’s ubiquitous WeChat. And US sanctions have deprived Chinese tech firms of the best chips to train their AI models, which could widen gaps between services like Ernie and their western counterparts.
In June, Baidu claimed Ernie outperformed ChatGPT on several measures. But Tencent said this month its own model in development was already one of China’s best, while Alibaba is integrating a ChatGPT-like AI into its meeting and messaging apps. All of them are seeking to make AI an integral part of their cloud-computing platforms, as such enterprise offerings are subject to less government scrutiny.
--With assistance from Sarah Zheng.
(Updates with share move in third paragraph)