Alphabet commits to US$75 billion capex plan amid rivals cooldown in AI infrastructure investment

Alphabet Inc. reported stellar results driven by AI advancements. Facing competition from ChatGPT, the company maintains its US$75 billion AI infrastructure investment despite rumors indicating Microsoft and AWS are slowing AI investments.

On April 24, 2025, Alphabet reported robust first-quarter results, with total revenue reaching US$90.23 billion, a 12% year-over-year increase. Google Services revenue grew 9.8% to US$77.26 billion, while Google Cloud surged 28% to US$12.26 billion. Operating income rose 20% to US$30.61 billion, with the operating margin expanding to 34% from 32%. Capital expenditures jumped to US$17.2 billion, up from US$12.01 billion, reflecting heavy AI and data center investments.

Alphabet's profit grew by 45.97% to US$34.54 billion. The company said in a filing that its first-quarter earnings included a US$8 billion unrealized gain from its investment in a private company's non-marketable equity securities. Bloomberg, citing an unnamed source, reported that the company is Elon Musk's Space Exploration Technologies Corp.

Sundar Pichai, Alphabet's CEO, said the strong results show healthy business growth fueled by their full-stack AI approach. This quarter featured the launch of Gemini 2.5, their most advanced AI model, marking a major innovation milestone. Search growth remained strong, supported by AI Overviews, which are now used by 1.5 billion monthly users. Paid subscriptions, led by YouTube and Google One, exceeded 270 million, while Cloud services saw rapid demand and growth.

Capex plan remains unchanged

Alphabet CFO Anat Ashkenazi outlined the company's capex outlook for 2025, expecting to invest around US$75 billion, with quarterly fluctuations due to delivery and construction timing. Increased capex has led to higher depreciation expenses, which rose 31% year-over-year in the first quarter, reflecting growth in technical infrastructure assets. This depreciation growth is anticipated to accelerate throughout 2025.

Bloomberg Intelligence said in a note that Alphabet reaffirmed its 2025 capital spending plan of US$75 billion, indicating strong demand for the company's datacom transceivers this year despite worries about slowing hyperscale AI investment. However, tariff-related risks could persist into 2026, potentially increasing input costs and impacting demand or margins.

Alphabet highlights AI offering growth and impact

Alphabet Inc. posted stronger-than-expected first-quarter earnings, boosted by robust search ad revenue and improving cloud profitability. Despite the benefits of AI startups using Google Cloud, the company faces stiff competition from AI tools like ChatGPT that challenge traditional search.

In the earnings call, Pichai extensively discussed the advancements and integration of the Gemini AI model, emphasizing its role as a cornerstone of Alphabet's AI strategy.

Pichai highlighted the rollout of Gemini 2.5, describing it as Alphabet's most intelligent AI model to date, with versions including 2.5 Pro and 2.5 Flash. Launched in the prior month, Gemini 2.5 Pro achieved top rankings on various benchmarks, excelling in reasoning, coding, science, and math, and debuted as the leading model in the chatbot arena. The 2.5 Flash variant was introduced to optimize cost and quality for developers, receiving positive feedback from both developers and consumers. Pichai noted a 200% growth in active users of AI Studio and the Gemini API since the year's start, underscoring strong adoption. Additionally, Gemini-based models like Imagine 3 and Veo 2 for image and video generation, and specialized models like Gemini robotics and AI Co-Scientist for health research, were cited as expanding Alphabet's AI capabilities into new domains.

Gemini models are now embedded in all 15 of Alphabet's products with over 500 million users, including Search, Android, Pixel, and Google Workspace. In Search, Gemini powers AI Overviews, serving 1.5 billion monthly users, and the experimental AI Mode, which handles complex queries.

Analysts probed Gemini's daily active users (DAUs), noting a reported 35 million DAUs compared to ChatGPT's larger base. Pichai emphasized a positive adoption cycle driven by recent model advancements and new features like Gemini Live, Deep Research, and Canvas, which have boosted engagement. He highlighted Gemini's differentiation through its integration across Alphabet's vast ecosystem, particularly in Search, where users engage deeply with AI-driven features.

According to Pichai, internally, Gemini enhances productivity, with over 30% of Google's checked-in code involving AI-suggested solutions, up from 25%, and is used in customer service and financial operations, including earnings call preparations. Pichai noted early agentic workflows in coding as a future growth area, with learnings being shared with Cloud customers.

Pichai expressed excitement about Gemini's roadmap, citing its state-of-the-art performance and Alphabet's organizational adjustments to capitalize on its momentum. While acknowledging competition, he underscored Gemini's unique positioning due to Alphabet's full-stack AI approach, from infrastructure (e.g., Ironwood TPUs and Nvidia GPUs) to world-class research and broad product integration. The focus remains on enhancing user experiences, driving developer adoption, and expanding enterprise applications, positioning Gemini as a key driver of Alphabet's long-term growth despite macroeconomic challenges like tariff headwinds.

Regarding AI Overview, Pichai highlighted its expansion to 15 languages and 140 countries, noting that users find Search to be more useful, boosting query diversity. AI Overviews maintain monetization rates comparable to traditional Search, providing a strong base for innovation, as stated by Chief Business Officer Philipp Schindler.

Source: Alphabet, April 2025

Source: Alphabet, April 2025

Source: Alphabet, April 2025

Article edited by Jerry Chen