On November 30, 2022, ChatGPT emerged unexpectedly, like a boulder dropped into a calm lake, sending ripples that drew numerous tech giants into the fray, thus kicking off the AI battle. After more than 900 days of intense competition and massive investment, this fierce battle has finally reached a critical turning point – several leading enterprises in the AI technology race have successfully converted their huge capital expenditures into substantial profit growth, and AI has begun to demonstrate its powerful “value-creating” capabilities.
In the latest round of the US stock market earnings season, tech giants such as Google, Microsoft, and Meta all delivered impressive “report cards”. Google’s parent company, Alphabet, reported second-quarter revenue of $96.428 billion, a year-on-year increase of 13.8%, and a net profit of $28.196 billion, a strong year-on-year growth of 19.4%, far exceeding market expectations. Microsoft’s fourth fiscal quarter revenue was $76.44 billion, a year-on-year increase of 18%, with its intelligent cloud business (including Azure) revenue reaching $29.88 billion, a significant year-on-year increase of 26%, both exceeding expectations. Its net profit was $27.2 billion, a year-on-year increase of 24%. Meta’s second-quarter revenue was $47.52 billion, a year-on-year increase of 22%, and its net profit was $18.34 billion, a year-on-year increase of 36%. All these eye-catching figures demonstrate the strong growth momentum that AI technology has brought to enterprises.
In fact, on the basis of having already invested heavily in AI, several tech giants have further increased their investment. Google’s capital expenditure in the second quarter was $22.446 billion, a year-on-year increase of 70%, and a quarter-on-quarter increase of over 30%. It also raised its full-year 2025 capital expenditure plan by $10 billion to $85 billion and expects to further increase it in 2026. Microsoft expects its capital expenditure in the first quarter of fiscal year 2026 to exceed $30 billion, a year-on-year increase of over 50%, far exceeding the previous analysts’ expectations of $24.23 billion. Meta’s full-year capital expenditure plan is between $66 billion and $72 billion, higher than the previous lower limit of the plan, and its capital expenditure in 2026 will also “increase significantly”.
These giants are spending so much money because AI computing power is still in short supply, and there is an urgent need to invest in AI infrastructure such as AI servers, networks, and data center construction and operation. In addition, Meta also needs to invest a large amount of money in talent recruitment to gain an advantage in the fierce talent competition. Previously, Zuckerberg poached a large number of people from various companies in Silicon Valley, stating that “infrastructure and talent will be Meta’s top priorities”, and employee compensation will also be the second-largest driver of cost growth for the company in 2026.
Looking back a year ago, the earnings reports released by tech giants were difficult to satisfy the market because the huge capital expenditures on AI did not bring equivalent actual returns. At the earnings conference, Google CEO Sundar Pichai was constantly questioned by analysts: Google invests $12 billion in AI every quarter, will it start to generate returns? Now, the giants can finally give a definite answer with their beautiful financial data and eye-catching user data: it was worth it!
Google’s Gemini app has 450 million monthly active users, and its daily call volume in the second quarter increased by 50% quarter-on-quarter. Gemini usage has soared 35 times year-on-year. The company output 480 trillion tokens per month in May, and this number increased to 980 trillion in June; the video generation model Veo3 has become a hit product, generating over 70 million videos since May. AI Overview has over 2 billion monthly active users (up from 1.5 billion in the previous quarter), and its monetization rate is roughly on par with traditional search. Several analysts have pointed out that AI has a positive impact on all of Google’s businesses, and Google’s investment in AI is “yielding comprehensive returns”.
In this earnings report and conference call, Microsoft disclosed two key pieces of data for the first time: one is the size of its Azure business in US dollars, with cloud service revenue from Azure and others expected to exceed $75 billion in fiscal year 2025, a year-on-year increase of 34%; the other is the monthly active user data for Copilot, with the total monthly active users of the Copilot series (including enterprise and consumer versions) reaching 100 million. Microsoft stated that the adoption of Microsoft 365 Copilot has increased the per-user revenue of Microsoft 365’s commercial cloud products (such as Office productivity software packages).
AI has also become a core engine driving business growth for Meta. Its operating margin in the second quarter was as high as 43%, and Zuckerberg emphasized at the earnings call that this was mainly due to the efficiency dividend brought by AI to the advertising system. Under the impetus of AI, the advertising conversion rate of Instagram increased by about 5%, and that of Facebook increased by 3%. AI creative tools are very popular among small advertisers. Additionally, the monthly active users of Meta AI have exceeded 1 billion and are penetrating into a wider range of scenarios.
According to GF Securities’ compilation, the monetization paths of the tech giants in AI are different but very clear: OpenAI leverages the first-mover advantage of its models to target high-end users; Google utilizes its unique full-stack advantage to drive the monetization of its product matrix; AWS focuses on the pay-as-you-go model (cloud services on demand + AI platform services and SaaS); Microsoft focuses on the platform (Azure cloud) + value-added subscriptions; Meta deeply empowers advertising + social media and reserves terminal businesses such as glasses.
Analysts point out that all the giants have seen the huge potential of AI in empowering traditional businesses, connecting existing businesses, and expanding new businesses. AI still has considerable value in areas such as intelligent cloud/agents/model API calls/software premium subscriptions/advertising and content monetization, which prompts the giants to accelerate their investments. Jim Tierney, the head of the US Growth Fund, stated that although capital expenditures still “seem endless”, given that the AI investments of these tech giants are showing returns, investors’ attitudes have completely changed. The positive market response to cloud computing revenue and AI-driven service sales is “truly remarkable”.
In the AI investment boom among tech giants, the term “FOMO” (fear of missing out) has been closely associated. Both Google and Meta have acknowledged this concern. Zuckerberg stated that once lagging behind in AI, “you will lose your position in the most important technology in the next 10-15 years”. Google CEO Sundar Pichai also believes that in the face of such a technological wave, “the risk of underinvestment is far greater than the risk of overinvestment”. As long as they can maintain a leading position, excessive capital expenditures are just a “small price to pay”.
When ChatGPT first emerged, the giants were driven by panic over technological disruption and rushed to enter the market and recruit, with the initial investment aimed at “securing a position”. But now, as the return path of AI investment becomes clearer and AI starts to truly “generate revenue”, the tech giants seem to be caught in an even more intense “FOMO 2.0”.
Take OpenAI as an example. Despite a significant increase in revenue this year, with an expected annual revenue of around $12 billion, the company is still plagued by high operating costs. Some analysis points out that for every dollar OpenAI earns, it incurs a cost of $2.25. A rough estimate suggests that the company’s operating expenses this year are expected to exceed $28 billion. To maintain its leading position, OpenAI still needs to raise a huge amount of funds. It is reported that the company is working on the $40 billion financing plan announced in March and has restarted the financing process on Monday, seeking funds from both new and existing investors.
Then there is Amazon. Despite several key figures in its latest financial report exceeding analysts’ expectations, the performance of its AWS business was “good but not outstanding”, dragging down its stock price. Some analysts directly described the sales growth of AWS as “very disappointing”, while others are concerned that it lags behind competitors in AI development due to the lack of a powerful AI model.
The Matthew effect in the AI industry is more pronounced than in the Internet era. Leaders form a “snowball effect” relying on the accumulation of data, computing power, and user habits, and the cost for latecomers to catch up increases exponentially. Moreover, model iteration, hardware upgrades, and ecological updates all require continuous heavy investment.
Statistics show that Google, Microsoft, Meta, and Amazon are expected to invest more than $350 billion in AI infrastructure such as data centers this year, and the figure will exceed $400 billion by 2026. “As long as revenue and order data exist, they can spend as they please,” said Brent Thill, an analyst at Jefferies. “This is an ongoing capital expenditure battle… probably only five companies are qualified to participate.”
The AI competition has entered a white-hot phase. Tech giants are constantly exploring and advancing with continuous investment, and the curtain of AI creating value has just risen.
