Tech giants fear falling behind, and none want to lag in the AI race.

These giants' apps have become super portals, penetrating multiple scenarios in our daily lives. Under the AI wave, these tech leaders are accelerating their embrace of AI, quietly preparing for the next competitive race.

Previously, comprehensive mainstream AI app rankings were compiled based on dimensions such as traffic and user reviews, selecting 100 notable global AI apps worth watching in 2025. Behind these AI products are not only emerging startups and unicorn companies, but also many listed enterprises.

Internet giants are simultaneously busy building their own AI apps while increasing investments in AI startups, pursuing dual-track AI deployment.

For example, Alibaba has not only launched products like Tongyi series and Quark that made it onto the top 100 AI app list, but also invested in 9 AI startups, including Silicon Flow, Yi AI, MiniMax, and SenseTime, betting on other promising AI products in parallel.

This raises the question: Is the AI business really that "attractive," and is it truly worth companies going all-in?

**BAT Invests Heavily in AI, Burning Approximately 80 Billion Yuan in Half a Year**

To explore this question, data was collected from current internet companies listed on A-shares and Hong Kong stocks, identifying over ten companies that mentioned AI business, investments, and profitability situations, attempting to glimpse answers from their 2025 half-year reports.

Among these listed internet companies, well-funded leading giants, especially the BAT camp, continue choosing to invest heavily in AI layout.

In just the first half of this year, the three BAT companies burned approximately 81.73 billion yuan on AI. Specifically, Baidu invested 8.83 billion yuan in R&D expenses for core businesses; Tencent used 22.9 billion yuan in capital expenditure to support AI business; regarding Alibaba, CEO Eddie Wu mentioned that Alibaba invested over 10 billion yuan in AI infrastructure and product development over the past four quarters.

From R&D expenses to capital expenditure, BAT's continuous AI investment targets both future AI technology breakthroughs and quietly accumulates current and future AI assets, equipping themselves for the AI competition.

Generally speaking, to assess whether a listed company takes R&D seriously enough, the proportion of R&D investment to total revenue is a key indicator. This reflects whether companies are willing to reinvest a significant portion of their revenue back into technology R&D, using AI for long-term investment.

From this perspective, companies like SenseTime, Kingsoft, and Wondershare's AI resource allocation is also considerable, with AI investment ratios all exceeding 30%.

SenseTime, one of the "AI Four Dragons," has directly gone "all-in on AI." Data shows that SenseTime invested nearly 90% of its revenue in AI-related R&D during the first half of 2025, with expenditure reaching 2.119 billion yuan, a 12% increase from the same period last year. This cost increase mainly stems from SenseTime's intensified efforts in foundation model training, fine-tuning, and generative AI application development.

**Listed Companies' AI Business Landscapes Show Diverse Approaches**

Indeed, AI has become a universal choice in the internet industry. Whether BAT-type super companies or enterprises focusing on niche sectors, all are finding ways to integrate AI into their business systems.

This can be directly observed from various listed companies' business layouts. Based on detailed business segments and descriptions disclosed in financial reports, we dissected the AI usage logic in corresponding businesses and organized AI's actual role in each company's operations.

Data shows that among these internet listed companies' businesses, AI-dominant businesses don't account for a high proportion; more business scenarios use AI as an enabling tool, embedding it into their products or business processes.

Taking two leading giants, Alibaba and Tencent, as examples, both have made cloud services their core AI layout direction in recent years (belonging to Cloud Intelligence and Financial Technology & Enterprise Services segments respectively). They currently have preliminary scale but still have considerable gaps compared to some of their other business lines.

Kunlun Tech, which deeply cultivates digital creativity, follows a similar pattern. On one hand, its three revenue pillar businesses - advertising, Opera search, and overseas social networks - are accelerating AI integration for upgrades. On the other hand, its newly launched AI software technology business in 2024, while currently having limited revenue scale, already represents 1.67 times last year's full-year amount, showing strong growth momentum.

However, some enterprises have carved out AI-centric paths, like iFlytek, SenseTime, and Wondershare, where AI has clearly become their core business.

Take iFlytek as an example: its smart education, smart city, smart automotive, and smart healthcare segments are all deeply integrated with AI. Especially its smart education business, leveraging products like AI blackboards and AI learning devices, accounted for over 30% of total revenue in the first half of 2025.

**Kuaishou's Keling AI Earns 2.5 Billion Yuan in Half Year - How High is Tech Giants' AI "Gold Content"?**

However, it's worth noting that listed companies' high-investment, wide-net AI deployment models don't necessarily bring proportional high returns.

In 2024, Kunlun Tech experienced its first annual loss since going public in 2015, and this loss trend wasn't curbed in the first half of 2025, with current attributable net profit losses of approximately 856 million yuan, further increasing year-over-year.

This is precisely due to high R&D expenses, as AI business has always been known for "burning money." Although its new AI business has shown initial growth speed, its current scale only accounts for 1.7% of overall business, far from reaching cost-balancing levels.

So, with companies burning so much money, who actually made money?

In the past half year, companies that could genuinely state their AI business "gold content" in financial reports weren't numerous.

From financial reports, companies' AI monetization methods mainly present two models: one relies on AI for direct revenue increase, while the other uses AI as a supporting role for indirect revenue increase.

Kuaishou represents the former model. Kuaishou's Keling AI is a leader among domestic and international video generation tools. On August 27, in A16z's latest global AI top 100 ranking, Keling AI's web version ranked even higher than overseas competitors like Sora and Runway.

Kuaishou's financial report shows that Keling AI not only made money but increasingly so. In Q2 2025, Keling AI generated 2.5 billion yuan, 1 billion yuan more than the previous quarter's disclosed figure, with commercial monetization accelerating.

Similarly, Kingsoft and 360 performed well in vertical sectors. The former drove office software business revenue growth through WPS's rich AI functions, while the latter achieved year-over-year revenue growth through "AI+Security" and "AI+Search" deployments.

Currently, many companies are focusing on vertical scenarios, exploring AI commercialization paths that both align with their business attributes and offer greater certainty.

On the other hand, current financial reports also show that not all enterprises' various businesses mention direct benefits from AI. Combined with the above business layout analysis, for some enterprises, AI functions more like an auxiliary tool for existing businesses.

For Tencent and Bilibili, AI's revenue boost is also prominent in advertising and marketing businesses. However, this boost isn't directly from AI but through AI's optimization of advertising creation, placement, recommendation, and effect analysis processes, thereby improving ad click-through rates and conversion rates, ultimately driving advertising business revenue growth.

In 2023, Tencent's President Martin Lau stated that AI technology had already brought considerable revenue to Tencent's existing businesses, but short-term benefits were more concentrated in advertising. By Q2 this year, he revealed that AI had contributed to Tencent's existing business growth and monetization in different ways, with potential future exploration of more viable To C AI monetization approaches.

In past years, when large companies were frantically burning money on AI layout, making money from AI seemed like a distant story. But now, the situation appears to have quietly changed.

However, even though AI can currently generate revenue, compared to tech giants' massive investments, these companies achieving profitability and positive cycles solely through AI itself still seems some distance away.