CITIC Group will strive to come up with over 50 in-depth applications powered by artificial intelligence to nurture specific foundation models for different industries, said the State-owned financial conglomerate during the four-day 2024 World AI Conference underway in Shanghai since Thursday.

This is part of CITIC Group's "AI+" action plan released on Friday.

Specifically, CITIC's financial service business units will focus on the AI applications in intelligent customer service, wealth management, and intelligent risk control, Wang Guoquan, the group's vice-president, said on Friday.

The group's advanced intelligent manufacturing and advanced materials sectors will focus on the AI application in industrial vision and automation technology. Efforts will be made to seek breakthroughs in applying AI in publishing, internet security and urban management, said Wang.

Meanwhile, the group will moderately advance the construction of its computing network, he said.

CITIC will build green, intelligent and large-capacity computing power in the western part of China, mainly to serve the research and development of foundation models. It will also construct efficient and intensive computing power in the eastern part of the country to explore more business application scenarios, said Wang.

Also on Friday, the group released three major achievements it has made in the development of foundation models.

Bond Copilot, developed by CITIC Securities, is one of them. It can integrate business opportunity mining, risk management, smart office services and project operation in the bond business sector.

The other two AI foundation models were released by China CITIC Bank and CITIC Press Group. The former provides a generative AI-based one-stop solution for corporate clients, while the latter creates a platform covering all the production processes in the publishing industry.

During a speech made on Friday to depict the AI applications outlook, Wu Chaoze, head of the research institute for China Securities, said the application of AI foundation models in the financial sector will mature in the next three to four years, while such application is only for simple non-decision-making processes at present.

But AI application has already significantly reduced copper exploration time in overseas markets. Foundation models have demonstrated much potential in terms of pharmaceutical R&D and health management, said Wu.

In the industrial sector, foundation models can optimize the entire production flow and decision-making mechanism. But high-quality data and accuracy are the main limitations of AI applications in the industrial field, she added.