China Accelerates AI Revolution: Approves 14 Industry-Specific Large Language Models (LLMs), Including Xiaomi, 01.AI, and Frontis.AI, Shaping a New Era in AI Development

China approves 14 large language models and enterprise applications for commercial use. Industry-specific LLMs reflect the push for efficiency in enterprises. Frontis.AI, ThreatBook, and XinYi Tech among the companies granted approvals.

Credit: Shutterstock

In a move to drive the adoption of artificial intelligence (AI) across various industries, China has given the green light to the first batch of large language models (LLMs) this year. These LLMs, which are the technology behind ChatGPT and similar services, have been approved for commercial use, signaling the country's commitment to AI development.

According to local media reports and company announcements, a total of 14 LLMs and enterprise applications have received approval. Among the companies granted permission are smartphone giant Xiaomi, AI specialist Beijing Fourth Paradigm Technology Co, and tech unicorn 01.AI, founded by venture capitalist Lee Kai-fu.

This latest batch of approvals includes industry-specific LLMs, reflecting the focus on boosting efficiency in enterprises. It marks a shift from previous approvals that mainly covered general AI models. The development and regulation of AI in mainland China are driving the adoption of technology tailored to specific industries.

These recent approvals follow the release of a number of generative AI services to the public in August last year. In anticipation of further advancements, the China Electronic Standardisation Institute was established to implement a national standard for LLMs.

Frontis.AI, a startup founded by former JD.com senior vice-president Zhou Bowen, received approvals in January for its "Pinshang" and "Moxiaoxian" LLMs. These models provide consumer product companies with industry insights, as well as support for product and marketing innovation. Frontis.AI plans to offer more customized services to its enterprise clients.

In addition to Frontis.AI, industry-specific LLMs from cybersecurity firm ThreatBook and video solutions provider XinYi Tech have also been approved. ThreatBook's XGPT AI model focuses on internet security, providing enterprises with timely and accurate analysis of potential threats. XinYi Tech's LLM, on the other hand, offers AI-powered video-generating and processing tools for areas such as advertising, education, media, and e-commerce.

With the latest approvals, the number of government-approved LLMs and related applications in China now exceeds 40, highlighting the country's commitment to AI development.

However, not everyone is convinced of the value of the LLM frenzy. Robin Li Yanhong, the founder, chairman, and CEO of Baidu, one of China's leading internet search and AI companies, criticized the excessive focus on LLMs. In a speech last November, Li called it a "huge waste of resources" and urged the government to prioritize the development of application initiatives that would bring more ChatGPT-like services to the market. He noted that there were already 238 LLMs introduced in the country as of October, most of which are still awaiting government approval.

---

• China approves 14 large language models and enterprise applications for commercial use

• Industry-specific LLMs reflect the push for efficiency in enterprises

• Frontis.AI, ThreatBook, and XinYi Tech among the companies granted approvals

Source: SCMP