China's public sector accelerates AI adoption in 2024 as Zhipu and iFlyTek emerge as winners

Large Chinese firms have accelerated their pace of adopting artificial intelligence (AI) this year, with the number of related contracts surging in the first half, government data show.

The number of tender contracts for services involving the use of large language models (LLMs) that found successful bidders more than doubled between the first and second quarters - from 23 in the three months through March to 58 from April to June 24 - according to figures published on the China Government Procurement and China Tendering and Bidding Public Service Platform websites. The contracts all used the keyword damoxing, the Chinese term for LLM.

LLMs are the technology underpinning conversational AI bots such as OpenAI's ChatGPT. Since Microsoft-backed OpenAI debuted its hit chatbot in late 2022, Chinese technology firms have raced to put out hundreds of their own LLMs and products that run on them. Only 1 tender contract related to LLMs was signed in the first half of 2023, before a jump in the last quarter of the year, the websites show.

Do you have questions about the biggest topics and trends from around the world? Get the answers with SCMP Knowledge, our new platform of curated content with explainers, FAQs, analyses and infographics brought to you by our award-winning team.

The contracts reflect interest from large Chinese firms, as public disclosures are only required for projects that either involve public interest and safety, public funds or loans from foreign entities. Tech giants Baidu, Huawei Technologies and Tencent Holdings - in addition to some well-funded start-ups - were among the providers with successful bids in the first half.

The numbers offer a glimpse into the growing adoption of AI in China, who is gaining the most traction in this space, and the industries in which these technologies are being applied.

Operators in energy, telecoms, finance and scientific research are the most eager to tap into the potential of LLMs, with these industries seeing 19, 14, 12 and 10 deals, respectively.

A district environmental protection agency in Beijing bought bespoke LLMs to help it predict flood seasons. The Nuclear Power Institute of China and China Merchants Securities are counting on LLMs developed by Beijing-based start-up Zhipu AI to help organise knowledge from materials accumulated over years of operations for more convenient staff use, according to contracts reviewed by the Post.

Chinese energy powerhouses are also looking to use LLMs to find deficiencies across a variety of equipment used in power grids and for oil exploration.

The most prolific bidders were not China's traditional internet giants. Zhipu AI, known as one of China's four "AI Tigers", has won the most LLM-related contracts this year at 12. AI firm iFlyTek, a US-sanctioned company known for its voice-recognition technology, successfully bid on 10 contracts from Chinese state-owned enterprises and government agencies.

Baidu and Huawei followed with five and three successful bids, respectively.

From month to month, there was a significant pickup in the pace of LLM contract bidding in the second quarter. January, February and March saw nine, five and nine contracts with successful bids, respectively. In April, that number jumped to 20 and then 24 in May. There have been 14 in June.

Out of the 81 contracts with successful bids this year, the most lucrative came from PipeChina, officially known as China Oil & Gas Piping Network Corporation, which is paying 152.6 million yuan (US$21 million) for LLM-related tech from Taifu Industry, a subsidiary of state-affiliated Shandong Energy Group.

Taifu will develop both software and hardware for training and developing LLMs for PipeChina, which oversees the management of the nation's extensive oil and gas pipelines. The contract did not specify how those LLMs will be used.

The total value of all 81 contracts is 433 million yuan, with an average of 5.3 million yuan per contract.

This article originally appeared in the South China Morning Post (SCMP), the most authoritative voice reporting on China and Asia for more than a century. For more SCMP stories, please explore the SCMP app or visit the SCMP's Facebook and Twitter pages. Copyright © 2024 South China Morning Post Publishers Ltd. All rights reserved.

Copyright (c) 2024. South China Morning Post Publishers Ltd. All rights reserved.