--- title: "Haitong Securities: AI accompanies rapid market growth, emotional large model scenarios open up" type: "News" locale: "en" url: "https://longbridge.com/en/news/227168845.md" description: "Haitong Securities released a research report indicating that the AI companionship market is experiencing rapid growth, and emotional large models will open up the human-machine companionship market in the next 2-3 years. Emotional computing and multimodal AI provide the technological foundation for emotional large models, driving the development of industries such as customer service, education, healthcare, and entertainment. The number of monthly active users on AI companion platforms has increased 30 times since 2018, and the market size is expected to grow from USD 30 million in 2023 to USD 15 billion in 2030, with an average annual compound growth rate of 236%" datetime: "2025-02-05T07:19:03.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/227168845.md) - [en](https://longbridge.com/en/news/227168845.md) - [zh-HK](https://longbridge.com/zh-HK/news/227168845.md) --- # Haitong Securities: AI accompanies rapid market growth, emotional large model scenarios open up According to the Zhitong Finance APP, Haitong Securities released a research report stating that in recent years, the AI companionship market has experienced rapid growth, and emotional intelligence has become a new frontier in the AI field. For the C-end, a large number of AI companionship applications represented by Character.AI and Poly.AI have blossomed and become popular choices in the AI application market; for the B-end, emotional large models empower enterprises in various industries and fields such as customer service, education, healthcare, and entertainment, with large models that combine EQ and IQ expected to open up the human-machine companionship market in the next 2-3 years. ## **Haitong Securities' main viewpoints are as follows:** **Emotional computing and multimodal AI build the technical foundation for emotional large models** Emotional computing refers to the use of computer technology to recognize, understand, and simulate human emotions, including three core tasks: emotion recognition, emotion content generation, and emotion response. With the support of general large models, emotional computing technology has made significant progress, demonstrating outstanding performance in emotion recognition, generation, and response, leading to the emergence of emotional large models. In addition, the development of multimodal AI has enabled large models to effectively capture and process multimodal information such as facial expressions, voice, text, and physiological signals, making emotion recognition and analysis more accurate and comprehensive. However, how to effectively address the long-term memory issue remains a technical challenge for emotional large models. **AI companionship applications flourish, opening up scenarios for emotional large models** In recent years, the AI companionship market has seen rapid growth. According to data from ARK Investment Management LLC cited by the extraordinary production research Tencent account, from 2018 to 2023, the number of monthly active users of AI companion platforms increased 30-fold, and the global market size for AI companions is expected to grow from $30 million in 2023 to $150 billion by 2030, with an annual compound growth rate of 236%. For the C-end, a large number of AI companionship applications represented by Character.AI and Poly.AI have blossomed and become popular choices in the AI application market. According to data from a16z cited by the Newin Tencent account, the user interaction frequency of companionship AI applications far exceeds that of other types of AI applications. At the same time, AI pet robots that can provide emotional value are gradually entering people's view, with a similarly vast market space. For the B-end, emotional large model vendors can provide API calls, AI dialogue solutions, and other services to enterprises in different industries, empowering customer service, education, healthcare, entertainment, and other sectors with emotional large models. **Typical cases and progress** Currently, emotional large models can be divided into two iterative routes: one is to iterate multimodal emotional computing capabilities based on general large models, thus giving rise to AI emotional applications, such as the iteration of GPT to GPT-4o and ByteDance's launch of the Doubao real-time voice large model; the other is to launch AI products based on generative large models focused on the emotional field, such as AEGMV from Zhuyv Future Technology and EVI from Hume AI. **Investment recommendations:** **AI Applications:** Kingsoft Office (688111.SH), Rainbow Soft Technology (688088.SH), Hehe Information (688615.SH), iFLYTEK (002230.SZ), Wanjia Technology (300624.SZ), New Wisdom Software (688590.SH), Hand Information (300170.SZ), New Guodu (300130.SZ), Danghong Technology (688039.SH), Tonghuashun (300033.SZ), Foxit Software (688095.SH), Fanwei Network (603039.SH), Zhiyuan Interconnection (688369.SH), 360 (601360.SH), SenseTime-W (00020), Maifushi (02556), Dingjie Smart (300378.SZ), Saiyi Information (300687.SZ), China Software International (00354), Hikvision (002415.SZ), Dahua Technology (002236.SZ); **AI Computing Power:** Runze Technology (300442.SZ), Inspur Information (000977.SZ), Huqin Technology (603296.SH), Zhongke Shuguang (603019.SH), Haiguang Information (688041.SH), Cambrian-U (688256.SH). **Risk Warning:** Industry policy risk, intensified market competition risk, AI technology development not meeting expectations, AI application implementation not meeting expectations ### Related Stocks - [01833.HK](https://longbridge.com/en/quote/01833.HK.md) - [300288.CN](https://longbridge.com/en/quote/300288.CN.md) - [002432.CN](https://longbridge.com/en/quote/002432.CN.md) - [00241.HK](https://longbridge.com/en/quote/00241.HK.md) - [BYTED.NA](https://longbridge.com/en/quote/BYTED.NA.md) - [300442.CN](https://longbridge.com/en/quote/300442.CN.md) - [00777.HK](https://longbridge.com/en/quote/00777.HK.md) - [002230.CN](https://longbridge.com/en/quote/002230.CN.md) - [002955.CN](https://longbridge.com/en/quote/002955.CN.md) - [002599.CN](https://longbridge.com/en/quote/002599.CN.md) - [300253.CN](https://longbridge.com/en/quote/300253.CN.md) - [300130.CN](https://longbridge.com/en/quote/300130.CN.md) - [AI.US](https://longbridge.com/en/quote/AI.US.md) - [300687.CN](https://longbridge.com/en/quote/300687.CN.md) - [09901.HK](https://longbridge.com/en/quote/09901.HK.md) - [603296.CN](https://longbridge.com/en/quote/603296.CN.md) - [688369.CN](https://longbridge.com/en/quote/688369.CN.md) - [01698.HK](https://longbridge.com/en/quote/01698.HK.md) - [603039.CN](https://longbridge.com/en/quote/603039.CN.md) - [688039.CN](https://longbridge.com/en/quote/688039.CN.md) - [688507.CN](https://longbridge.com/en/quote/688507.CN.md) - [00700.HK](https://longbridge.com/en/quote/00700.HK.md) - [300170.CN](https://longbridge.com/en/quote/300170.CN.md) - [300494.CN](https://longbridge.com/en/quote/300494.CN.md) - [601360.CN](https://longbridge.com/en/quote/601360.CN.md) - [02556.HK](https://longbridge.com/en/quote/02556.HK.md) - [688615.CN](https://longbridge.com/en/quote/688615.CN.md) - [688095.CN](https://longbridge.com/en/quote/688095.CN.md) - [002236.CN](https://longbridge.com/en/quote/002236.CN.md) - [00354.HK](https://longbridge.com/en/quote/00354.HK.md) - [603019.CN](https://longbridge.com/en/quote/603019.CN.md) - [301252.CN](https://longbridge.com/en/quote/301252.CN.md) - [688590.CN](https://longbridge.com/en/quote/688590.CN.md) - [688111.CN](https://longbridge.com/en/quote/688111.CN.md) - [01896.HK](https://longbridge.com/en/quote/01896.HK.md) - [688088.CN](https://longbridge.com/en/quote/688088.CN.md) - [300624.CN](https://longbridge.com/en/quote/300624.CN.md) - [002415.CN](https://longbridge.com/en/quote/002415.CN.md) - [00020.HK](https://longbridge.com/en/quote/00020.HK.md) - [688603.CN](https://longbridge.com/en/quote/688603.CN.md) - [DAO.US](https://longbridge.com/en/quote/DAO.US.md) - [000977.CN](https://longbridge.com/en/quote/000977.CN.md) - [OpenAI.NA](https://longbridge.com/en/quote/OpenAI.NA.md) ## Related News & Research - [CICC Remains a Buy on Alibaba Health Information Technology (ALBHF)](https://longbridge.com/en/news/286663292.md) - [CICC Reaffirms Their Buy Rating on Ping An Healthcare and Technology Company (PANHF)](https://longbridge.com/en/news/286634736.md) - [Alibaba Health Declares Final Dividend for FY2026](https://longbridge.com/en/news/286406079.md) - [15:45 ET"Lean" Strengthens Saudi Arabia's Presence in Global Digital Health and Precision Medicine Ecosystems](https://longbridge.com/en/news/286811787.md) - [05:54 ETTermHub™ Launches New Managed FHIR Offering for SDOs](https://longbridge.com/en/news/287042225.md)