💰 Google's parent company just crossed $400 billion in annual revenue for the first time ever. The driving force? AI. Cloud revenue surged 48%, the Gemini AI assistant hit 750 million monthly users, and the company plans to pour up to $185 billion into AI infrastructure in 2026 alone. How is this reshaping Japan — the world's third-largest economy — and its approach to AI?
Alphabet Shatters the $400 Billion Revenue Barrier
On February 4, 2026, Alphabet Inc. — Google's parent company — reported its Q4 2025 earnings, blowing past analyst expectations and marking a historic milestone. For the first time, the company's full-year revenue exceeded $400 billion, reaching $402.8 billion (approximately ¥63 trillion).
Q4 revenue alone climbed 18% year-over-year to $113.8 billion, while net income surged nearly 30% to $34.46 billion. Earnings per share came in at $2.82, well above the $2.63 analysts had predicted.
CEO Sundar Pichai set the tone in the earnings call: "The launch of Gemini 3 was a major milestone and we have great momentum."
AI Is Driving Everything
The most striking feature of these results is how thoroughly AI has permeated every revenue stream.
Google Cloud was the star performer, with Q4 revenue jumping 48% year-over-year to $17.66 billion, pushing its annualized run rate above $70 billion. Even more impressive, Cloud's operating margin improved from 17.5% to 30.1% — proof that this isn't just growth, but profitable growth.
Behind these numbers is an explosion in enterprise AI demand. It's no longer just AI startups using Google Cloud. Major corporations across finance, manufacturing, and retail are integrating Google's AI services into their operations. The company's order backlog hit $240 billion, up 55% from the previous quarter, and the number of billion-dollar-plus deals in 2025 exceeded the previous three years combined.
Google's core search advertising business remained robust at $82.28 billion (up 13.5%), while YouTube ads grew 9% to $11.38 billion. YouTube's total annual revenue across advertising and subscriptions surpassed $60 billion — making it effectively one of the world's largest standalone media companies. Across all consumer services, Google now counts over 325 million paid subscribers.
Gemini: 750 Million Users and Counting
At the heart of Alphabet's AI strategy is Gemini, its family of AI models that compete directly with OpenAI's GPT series and Anthropic's Claude.
The latest iteration, Gemini 3, launched in November 2025, achieved state-of-the-art performance in reasoning and multimodal understanding — the ability to process text, images, video, and audio in an integrated way. Monthly active users of the Gemini app reached 750 million, up from 650 million just one quarter earlier. Gemini models now process over 10 billion tokens per minute through direct API use.
Perhaps most significantly, Alphabet slashed Gemini's serving costs by 78% over 2025 through model optimizations and efficiency improvements. This means higher-quality AI at dramatically lower cost — a competitive advantage that's hard for rivals to match.
The company also announced plans to sell its seventh-generation custom AI chip, the TPU "Ironwood," to cloud customers. The first major customer is Anthropic, which will receive up to one million units in a deal worth tens of billions of dollars.
The $185 Billion Bet on AI Infrastructure
The number that made investors gasp was the 2026 capital expenditure forecast: $175 billion to $185 billion. That's nearly double the 2025 spend and far above the $119.5 billion analysts had predicted.
To put this in perspective, $185 billion exceeds Japan's entire defense budget (approximately $57 billion for fiscal 2025) by more than three times. A single company is investing more in AI infrastructure than most nations spend on national security.
CFO Anat Ashkenazi explained that AI computing capacity will remain supply-constrained throughout 2026, making aggressive investment essential. With over $120 billion in cash reserves, Alphabet can afford to build what Ashkenazi described as the infrastructure for "meeting significant cloud customer demand."
The stock initially dipped about 3% in after-hours trading as investors digested the massive spending plan, before recovering. The message from leadership was clear: this isn't spending to survive — it's spending to win.
Waymo and the "Other Bets"
Not everything in Alphabet's portfolio is soaring. The "Other Bets" segment — which includes self-driving car company Waymo and life sciences unit Verily — reported a $3.6 billion loss for the quarter, up significantly from $1.17 billion a year ago.
However, the bulk of this increase came from a $2.1 billion stock-based compensation charge related to Waymo's new funding round at a $16 billion valuation. Waymo itself completed 15 million trips across five U.S. cities in 2025 and expanded to Miami in January 2026, showing operational momentum even as accounting charges weigh on the bottom line.
Google's Growing Footprint in Japan
Alphabet's AI expansion has direct implications for Japan.
Google opened its first data center in Japan in Inzai City, Chiba Prefecture in 2023, as part of a ¥100 billion ($650 million) "Digital Future Initiative" to support Japan's digital transformation. The company invested in infrastructure, digital skills training, and partnerships with local organizations. It also partnered with JERA, Japan's largest power company, through a 20-year virtual power purchase agreement to supply renewable energy to the facility.
In 2025, Google teamed up with KDDI, one of Japan's major telecom carriers, to operate Google AI services from a data center in Osaka. This directly addresses a major concern for Japanese enterprises: data sovereignty. Many Japanese companies have been reluctant to adopt cloud AI services due to fears about sensitive data leaving the country. Having domestic data centers eliminates that barrier.
In January 2026, Google launched the "Google AI Plus" plan in Japan, making advanced Gemini capabilities available at accessible price points for individual users. For businesses, Gemini AI assistants are now deeply integrated into Google Workspace tools like Gmail, Google Docs, and Google Meet, enabling AI-powered email drafting, document summarization, and meeting transcription in Japanese.
Japan's AI Adoption: Introduced but Underperforming
While Google's AI business booms globally, Japan's corporate AI adoption tells a more complicated story.
According to Japan's Ministry of Internal Affairs and Communications (MIC) 2025 White Paper, 55.2% of Japanese companies using AI report utilizing generative AI in some business capacity. But when measured by actual corporate adoption rates, Japan sits at just 27% — compared to 81.2% in China and 68.8% in the United States.
A 2025 survey by PwC Japan paints an even more nuanced picture. Japan's "AI promotion level" is roughly average among five surveyed nations (Japan, US, UK, Germany, China). But its "effectiveness level" — the actual business results companies achieve from AI — lags significantly behind.
In other words, Japanese companies are adopting AI, but they're struggling to turn that adoption into meaningful outcomes.
The structural reasons are revealing. Japanese companies tend to view AI primarily as a tool for "operational efficiency and filling labor shortages" (業務効率化や人員不足の解消). In contrast, US and UK companies see AI as a means for "business expansion and innovation." This fundamental difference in mindset shapes how AI gets deployed and what results it produces.
Additionally, about half of Japan's small and medium-sized enterprises (SMEs) have not yet established clear AI adoption policies. The combination of leadership gaps, AI talent shortages, and organizational inertia — "we can still do our jobs without AI" — creates significant resistance to transformation.
Japanese Companies Leading the Way
Not all Japanese companies are lagging behind. Several have achieved impressive results through committed AI strategies.
Panasonic Connect deployed a company-wide AI assistant called "ConnectAI" and reduced 448,000 hours of work annually in its second year of deployment. The key breakthrough was shifting from using AI as a "search tool" to delegating substantive tasks like code generation and document review.
Mitsubishi UFJ Bank rolled out ChatGPT access to 40,000 employees and estimates monthly labor savings exceeding 220,000 hours. The bank plans to invest approximately $3.3 billion in AI over three years.
Toyota is collaborating with Google Cloud on a hybrid cloud platform that enables manufacturing teams to develop AI models independently. Combined with a joint project with NTT to build a $3.3 billion "Mobility AI Platform," Toyota is positioning itself at the intersection of automotive and AI innovation.
What these leaders share is clear executive sponsorship for AI adoption, systematic deployment across the organization rather than isolated experiments, and a willingness to fundamentally rethink business processes rather than simply automating existing ones.
2026: From AI That Creates to AI That Acts
The broader AI industry is entering a pivotal transition in 2026. The dominant paradigm is shifting from "generative AI" — which produces text, images, and code — to "agentic AI" — which autonomously plans, decides, and executes multi-step tasks across systems.
Google is at the forefront of this shift. Its development platform "Google Antigravity" already has over 1.5 million weekly active users building AI agents that can autonomously plan and execute complex software development tasks.
For Japan, this transition carries existential importance. The country faces a projected labor shortage of 11 million workers by 2040 due to its aging population and declining birthrate — a crisis often referred to as the "2040 Problem" (2040年問題). AI agents that can autonomously handle routine business processes aren't a luxury for Japan; they're a demographic necessity.
But benefiting from this technology requires building organizational readiness now. Companies need clear AI strategies from leadership, investment in AI literacy across all levels, and the courage to restructure workflows around AI capabilities rather than treating AI as an add-on to existing processes.
Alphabet's record-breaking earnings aren't just a financial story. They're a signal of how quickly AI is becoming the foundation of the global economy. For Japan, the question isn't whether to join this transformation — it's whether the pace of adaptation will be fast enough.
Japan is grappling with how to harness AI amid an aging society and shrinking workforce. How is AI adoption progressing in your country? Is the discussion focused on efficiency, innovation, or managing the disruption? We'd love to hear your perspective — share your thoughts in the comments below.
References
- https://blog.google/company-news/inside-google/message-ceo/alphabet-earnings-q4-2025/
- https://9to5google.com/2026/02/04/alphabet-q4-2025-earnings/
- https://www.cnbc.com/2026/02/04/alphabet-googl-q4-2025-earnings.html
- https://www.soumu.go.jp/johotsusintokei/whitepaper/ja/r07/html/nd112220.html
- https://www.pwc.com/jp/ja/knowledge/thoughtleadership/generative-ai-survey2025.html
- https://blog.google/intl/ja-jp/company-news/inside-google/data-center-in-inzai-city/
Reactions in Japan
63 trillion yen a year. That's over 10% of Japan's GDP ratio-wise. A single company operating at nation-state economic scale.
Stock dipped briefly on the 29 trillion yen capex plan, but they're sitting on 18 trillion yen cash. The financial muscle is in a different league. All Japanese IT companies combined can't match that.
78% reduction in Gemini inference costs is brutal. Great for small companies like ours accessing high-quality AI via API, but it raises a fundamental question: does building our own models even make sense anymore?
27% AI adoption rate? No way... hardly anyone around me uses AI. Big corporations inflate that number. SMEs are probably below 5%.
Power demand at the Inzai data center is getting insane. The AI era ultimately comes down to 'how do you secure electricity.' Can't separate this from the nuclear restart debate.
My office still uses fax machines and they talk about AI adoption... I just wish management would learn to use Excel properly first (seriously).
Google Cloud crossing 30% operating margin is historic. The three-way battle with AWS and Azure intensifies, but Google's TPU sales strategy gives them an edge. Japan's system integrator business model is being fundamentally disrupted.
750 million Gemini users means all that input data flows to Google, right? We need to seriously consider the risk of information concentration behind the convenience.
Students submitting AI-written reports is becoming the norm. Should education teach 'how to use AI' or 'how to think without AI'? We don't have the answer on the ground yet.
Working in a town of 30,000 people, AI transformed our citizen services. Multilingual counter AI is hugely appreciated by foreign residents. Delivering quality public services with small teams — that's AI's real value.
Every time I see 'AI will transform Japan' articles... our department workflow is identical to three years ago. Only change was getting Slack. This isn't even a DX problem — it's more fundamental.
Supporting AI projects at client companies. The difference between success and failure is crystal clear. Companies where the CEO says 'I created an AI department, it's their job' fail 100% of the time. Whether executives personally use and understand AI is the watershed moment.
Google Antigravity is legit. Feels like coding work dropped 30%. But I spend that freed time on new projects, so total work actually increased lol.
Honestly, with 10 years to retirement, I didn't care about AI. But seeing junior staff produce documents 3x faster made me worried. Age doesn't matter — you have to keep learning...
Is Japan okay just depending on Google and OpenAI? We should support domestic AI like NTT's tsuzumi and Preferred Networks more. From a national security perspective, having foreign companies control core technology is risky.
Hooked on generating images with Gemini. Can make presentation visuals with it now, so honestly, illustration commissions dropped dramatically. Feel bad, but this is reality...
Google's $185B capex exceeds many US states' annual budgets. If AI infrastructure is the new 'digital national power,' Japan needs to compete or risk being left behind.
China has Baidu, Alibaba, and Tencent building their own AI ecosystems. Why hasn't Japan cultivated its own tech giants? How long will they depend on American platforms?
In Sweden, even SMEs are adopting AI because the government invests heavily in digital skills education. Strange that Japan has the tech capability but low utilization. Education investment might be the key.
AI trading is standard in London's financial sector. Hearing Japanese banks just started rolling out ChatGPT to 40,000 employees felt like a time lag, honestly. But Panasonic cutting 448,000 hours is impressive.
AI startups pop up weekly in India. We have labor but need AI for global competitiveness. Japan has the opposite — labor shortages but not using AI. What a waste of potential.
German manufacturing also struggles with AI adoption, so Japan's situation isn't foreign to us. But Germany has spent 10 years on Industrie 4.0 factory digitalization. Japan should integrate AI more into its manufacturing floors.
In Brazil, Google's improved AI translation made Portuguese content much more accessible. If AI accuracy is improving for complex languages like Japanese too, Japanese companies should be leveraging it more.
France has active debates about the environmental impact of massive AI investments. Data center power consumption is exploding. Curious how Japan balances decarbonization with AI expansion.
Korea has Samsung and NAVER investing heavily in AI, but Google's scale is unmatched. The shared challenge for Japan and Korea: we're tech powerhouses with weak 'AI field implementation.' Hardware is strong but software lags.
UAE appointed an AI minister and pursues AI as national strategy, partnering with companies like Google at the government level. Japan's AI strategy looks overly cautious from our perspective.
In Poland's IT industry, Google's AI APIs are becoming the de facto standard. If it's cheap and high-performing, there's no reason not to use it. But single-vendor dependency risk should always be on your radar.
In Mexico, not speaking English meant missing out on AI benefits. Gemini's multilingual support is changing that. A Japanese friend says Japanese language quality improved dramatically. Breaking language barriers is wonderful.
Google's AI investment shocked Australia's startup scene. But what matters isn't being the one who builds AI — it's mastering how to use it. Japan should focus on building AI utilization skills.
Japanese engineer in Canada here. Almost zero resistance to AI at work here. When I visit Japan on business, I still hear 'AI can't be trusted' and 'who takes responsibility for errors?' — cultural barrier is real.
Nigeria still has areas with unstable internet, yet Google boasts 750M monthly AI users. The digital divide exists within advanced countries too — I imagine Japan's rural-urban gap is similar.