From computing power to applications, 2026 Capital is heavily investing in the new paradigm of AI marketing.

The commercialization of AI applications and how to achieve scale in business monetization are currently hot topics in the capital market. In recent times, a series of landmark events have brought strong signals: Several domestic large model and AI application companies have successfully IPOed on the Hong Kong Stock Exchange, validating their business models and sustainable profit paths; Intelligent body start-up company Manus was acquired by Meta, Musk open-sourced the X platform recommendation algorithm, Wal-Mart integrated generative AI deeply into the shopping process, and other industry dynamics continue to ferment, allowing the secondary market to clearly see the huge potential for the outbreak of AI applicationsrelated concepts such as GEO (Generative Engine Optimization) / GEM (Generative Engine Marketing) have seen a rise; The most significant commercialization emphasis comes from OpenAI's key decision in January 2026: announcing advertising tests in the free version and Go version of ChatGPT. This is not only a revenue exploration for AI giants but also signifies a fundamental shift in how users access information, with a newly emerging market driven by generative AI accelerating. For the marketing industry, this means a deep restructuring of scenarios: as traffic entrance shifts from search boxes to dialog boxes, the entire industry's allocation mechanisms, technical architectures, and ecosystem roles will undergo a systematic change. The Era of AI Advertising Arrives For the average user, OpenAI's inclusion of advertising may seem like a minor adjustment in experience. When a user asks "recently, which wireless earphones are good, Beijing Zhidemai Technology," the system may naturally embed a recommendation in the reply, marked as "sponsored by the brand." However, from an industry perspective, while traditional internet users obtained information through keyword searches, the generative AI era is shifting towards natural language interactions. This surface-level change is triggering a deep-rooted innovation in global traffic distribution mechanisms. On one hand, there is a migration of traffic entry points and monetization channels. Data shows that as of December 2025, ChatGPT had 880 million monthly active users, Google AI Overviews covered over 2 billion users, while domestic apps such as Dou Bao, DeepSeek, and Qianwen each had over a billion active users. User behavior studies show that over 80% of users have developed a habit of obtaining consumption information through AI, with nearly 35% engaging in high-frequency interactions daily. AI has effectively become the new hub for traffic distribution, with this concentration of attention making brand placement in AI answers critical. Brands that do not appear in the "top answers" sequence signify lost competition opportunities for AI traffic entry. Adobe Analytics data shows that during "Black Friday" in 2025, AI-related traffic to US e-commerce websites increased by 600% year-on-year. This growth rate confirms the reality of traffic entry shifting. On the other hand, GEO/GEM brings about a revolution in the advertising marketing paradigm. From a technological standpoint, GEO essentially involves "precision intent capture," where traditional SEM relies on keyword bidding and brand rankings positions on search results pages; GEM, on the other hand, achieves natural integration of advertisements as answers within conversation contexts. OpenAI's advertising design confirms this logic: when a user inquires about the recipe for a Mexican dinner, ChatGPT provides the recipe and pushes links to purchase ingredients, even supporting follow-up questions like "can I substitute low-fat cheese." Industry testing data shows that this type of multi-round interactive advertising can increase conversion rates by over 3 times compared to traditional SEM. Unlike traditional SEO, which aims to improve webpage rankings, GEO's primary objective is to embed brand information into the interconnected search results of large models, ultimately becoming a part of AI responses. This demands that brand content not only be keyword-optimized but also deeply understand user intent based on AI and offer matching solutions at the right moments. In terms of technological requirements, GEO demands that brand content possess three key features: semantic recognizability (understandable by models), structural reliability (usable by models), and interactive persistence (value retention in multi-turn conversations). The rise in this technological threshold is giving birth to a new industry ecosystemservice providers that can connect brands with AI platforms, understand generative logic, and build delivery channels are becoming the infrastructure providers of the new era. The traditional SEO/SEM market size is approximately $1.04 trillion, giving rise to global digital advertising giants like Google, Meta, and Amazon. As AI brings about industry changes with GEO/GEM, market participants are now on the same starting line, where those with technological reserves and resource endowments have the potential to lead the new AI advertising era. AI Applications Enter Their Investment "Golden Year" The deeper significance of the ChatGPT advertising event lies in realizing the monetization of AI traffic and the industrial transformation of AI advertising marketing, marking the opportune moment for the scale commercial deployment and development of AI applications. This is not only a transformation for OpenAI but a landmark event marking the transition of the entire AI application industry from the "capital injection period" to the "business monetization period." Not only can consumer traffic be monetized, but in business applications, the explosive growth of AI marketing represented by GEO/GEM is on the horizon. Industry data supports this assessment. According to industry statistics, the global application market size exceeded $300 billion in 2025, with the Chinese market leading the world with a 45% growth rate. From an industry logic perspective, as competition intensifies in underlying large model technologies, it is application layer innovation that truly reaches users and generates commercial value. The development of GEM and GEO is a practical validation of the commercialization capabilities of AI applicationsthey not only demonstrate AI applications' ability to restructure traditional industry chains but also present a path for AI applications to create new value spaces. Specifically in the GEO industry, according to the "2025 China GEO Industry Development Report," the global GEO market size is expected to reach $24 billion by 2026, and is projected to reach $100 billion by 2030, with a compound annual growth rate of approximately 53%. From an investment perspective, AI applications have been highly sought after by the capital market at the beginning of the year. As leading large model companies in China, KNOWLEDGE ATLAS and MiniMax made successful debuts on the Hong Kong Stock Exchange in early January, receiving warm market reception and both surpassing a market value of over a hundred billion Hong Kong dollars, with the market bullish on AI application scenarios leading to a large influx of funds into related thematic ETFs, with gaming media ETFs showing growth of over 30%. Changjiang Research believes that AI applications are poised for a dual breakthrough in consumer and business sectors in 2026, recommending monitoring the movements of leading enterprises positioning themselves in various scenarios. Guosen indicates that the technology market is expected to transition from computing infrastructure to the application layer, acceleration in the commercialization of AI applications is anticipated. The logic behind the acceleration of AI applications is that as super traffic entrances emerge, user habits are shifting from "typing keywords to search for web pages" to "directly asking AI for answers." According to Gartner's forecast, by 2026, about 25% of global traditional search engine traffic is expected to transfer to AI tools, providing fertile ground for GEO growth. Furthermore, the accelerated increase in penetration rate is another key logic for investing in AI applications. With the announcement last year of the "Opinions on Deepening the Implementation of the 'Artificial Intelligence+' Action," it was proposed to achieve wide and deep integration of artificial intelligence with six major strategic areas by 2027, ensuring that new generation intelligent terminals and intelligent entities and other applications will have a popularity rate exceeding 70% by 2030, and over 90% beyond that. The explosive power of the GEO market stems from its ability to replace traditional marketing models and the rapid increase in AI search penetration rates. AI Advertising, China's Enterprise Technological Layout In this paradigm shift in AI advertising, the marketing industry has an opportunity for restructuring. However, the challenge lies in the fact that not all enterprises can comprehend the "new syntax" of GEM, let alone reconstruct the entire advertising system. Some Chinese enterprises have already completed advanced positioning, with leading AI marketing technology enterprise Tiandong Technology establishing a first-mover advantage in the GEM race through its multi-intelligent agent Navos Marketing Multi-Agent. Navos is positioned as an intelligent assistant truly able to understand marketing intent, comprehend the GEM structural grammar, and generate strategies, driven by Tiandong's self-developed "Titan Extreme" large model, possessing deep understanding capabilities of user intent, semantic context, and advertising structures. Specifically, Navos can automatically generate structured advertising material based on real-time marketing data and industry expertiseessentially generating high-quality materials based on industry popular cases and providing executable marketing strategies. Whether it's contextual tone analysis for different categories or content variations for multi-language markets, Navos can complete the entire chain from semantic understanding to strategy generation. It understands not just "how to advertise" but more importantly "how to be prioritized, called upon, and reached by users in the GEM context by large models". In practical terms, Navos employs natural language interaction with low operating thresholds. For example, when a user asks for "AI tools with multilingual support," Navos will output suitable "static card ads" suggestions and recommended dialogues tailored to ChatGPT contexts; When a user asks about "AI customer service vs traditional customer service, which is better," the system can generate a combination plan that caters to both long-form GEO and static materials. In the future, users will likely be able to generate advertising deployment strategies for search, social media, and AI applications through Navos with just a single sentence. This output not only lowers the creative threshold for enterprises entering the GEM scene for the first time but also establishes an executable loop between materials and strategies, significantly enhancing deployment efficiency and operational capabilities. Supporting all of this is Tiandong's years of accumulated real marketing data and frontline experience on a global scale. By 2025, Tiandong had served over 100,000 advertisers, with business coverage spanning over 200 countries and regions globally, with a customer coverage rate exceeding 80% of China's outbound major clients, including top enterprises such as Alibaba, TikTok, Temu, Ku Luo Games, Mi Ha You, and others. More importantly, this data and customer base can naturally migrate to the GEM scene. The core requirement of GEM advertising is to precisely deliver content to users most relevant in semantics, and across global mainstream media platforms, Tiandong has supported hundreds of millions of advertising material tests, A/B experiments, and user audience matching, resulting in a real, usable, structurally clear, and highly scene-specific marketing data asset. This comprehensive ability of "technology + data + ecosystem" is specifically reflected in Tiandong's expertise in deciphering mainstream AI platform algorithms, reverse-attribution insights, rapid productification capabilities, and customer stickiness established through market positioning. These elements constitute the core competitive advantage of enterprises in the GEM era. For the capital market, this means that the investment logic is shifting from "technological concepts" to "application implementation." Those enterprises that first understand AI application scenarios, establish technological capabilities, and complete ecosystem layouts are standing at the starting point of a new industrial cycle.