Over the past two years, global tech giants have engaged in an unprecedented race to invest in artificial intelligence. Companies like Microsoft, Google, Amazon, and Meta have collectively poured hundreds of billions of dollars into building AI infrastructure, with NVIDIA emerging as the primary beneficiary of this cycle thanks to its GPUs.
However, as we move into 2026, the market’s focus is undergoing a fundamental shift.
Previously, investors asked, "Who is investing the most in AI?" Now, the real question is, "Who can generate genuine revenue and profits from AI?" The answer to this question is redefining how AI stocks are valued.
The AI Value Chain Enters the Revenue Verification Phase: Investment Logic Is Being Reshaped
Data analysis from research firm Exponential View, covering more than 1,000 companies worldwide, shows that in Q1 2026, global (excluding China) quarterly AI industry revenue surpassed the corresponding infrastructure depreciation costs for the first time. Specifically, the combined AI-related sales revenue of hyperscale and emerging cloud service providers reached around $25 billion, marking the second consecutive quarter that this figure exceeded the estimated depreciation cost of AI data centers and chip investments, which stood at $21 billion. As of June 2026, global annualized revenue from generative AI hit $175 billion (actual revenue for the past 12 months was about $110 billion).
This milestone means the AI industry has crossed its first threshold for "self-sustaining growth"—current AI operations generate enough cash flow to cover the accounting depreciation costs of servers, GPUs, and data centers. This marks a pivotal moment: AI has moved from the "cash burn phase" into the "self-verification phase."
Looking at the value chain, companies at different stages are experiencing distinct commercialization cycles:
AI Chip Segment: NVIDIA, AMD, and Broadcom are the core players, benefiting from ongoing growth in computing power demand.
Cloud Infrastructure Segment: Microsoft, Google, and Amazon are converting AI service revenue into commercial success.
AI Server Segment: Dell and Super Micro Computer profit from data center expansion.
Semiconductor Equipment Segment: ASML and Applied Materials support chip manufacturing needs.
AI Software Segment: Adobe and Salesforce are driving enterprise application commercialization.
NVIDIA: The Biggest Winner of the AI Infrastructure Era?
NVIDIA (NVDA)
NVIDIA remains the most central company in the AI value chain. For fiscal year 2026, NVIDIA’s total revenue reached $216 billion, up 65% year-over-year. Data center business revenue hit $194 billion, up 68%. Operating cash flow reached $103 billion. Analysts’ average price target for NVIDIA is about $558.77.
Previously, the market traded on the logic that "AI needs GPUs." Going forward, the key question is: Can GPU revenue continue to translate into high profit growth? At the 2026 shareholder meeting, NVIDIA CEO Jensen Huang stated that the question of AI investment returns "already has an answer." He further emphasized that higher token throughput leads to higher revenue—a logic that underpins NVIDIA’s financial performance.
However, risks are also mounting: AMD’s MI series AI chips are vying for data center market share; cloud providers (such as Google’s TPU, Amazon’s Trainium and Inferentia) are accelerating deployment of their own chips; custom ASIC chips are also encroaching on traditional GPU market space. The AI chip market is shifting from "NVIDIA’s sole dominance" to a diversified landscape of "GPUs + ASICs + specialized chips."
Microsoft: The Benchmark for AI Commercialization
Microsoft (MSFT)
Microsoft boasts the Azure cloud platform, strategic partnership with OpenAI, and the Copilot ecosystem, forming a complete loop for AI commercialization.
In Q3 FY2026, Microsoft’s revenue reached $82.9 billion, up 18% year-over-year. Annualized AI business revenue hit $37 billion, surging 123% year-over-year, making Microsoft the clear leader in enterprise AI transformation. Azure and other cloud services revenue grew 40% year-over-year, with company guidance for the next quarter at constant currency growth of 39%-40%, noting that customer demand still exceeds available capacity. Analysts’ average price target is $558.77.
Microsoft faces a key challenge: Will enterprises continue to pay for AI productivity tools? Deutsche Bank estimates that Azure’s actual gross margin dropped from about 60% in FY2024 to 52% in FY2026, with incremental gross margin hovering around 40% for seven consecutive quarters. However, AI service gross margins are recovering from the 25%-30% range in FY2026 to above 30%, and Azure’s gross margin is expected to stabilize at 50%-55% in FY2027/2028.
Microsoft projects capital expenditures of about $190 billion for 2026. Balancing massive capital spending with AI revenue growth will be the key variable determining Microsoft’s valuation trajectory.
Google: Can Gemini Reshape the AI Competitive Landscape?
Alphabet (GOOGL)
Google’s Gemini large model, Google Cloud platform, and the world’s largest search ecosystem form the three pillars of its AI strategy.
In Q1 2026, Google Cloud revenue reached $20.03 billion, up 63% year-over-year—far exceeding analysts’ expectations of about $18 billion. Cloud business operating profit jumped from 17.8% a year ago to 32.9%, indicating not only scale growth but significantly improved profitability. Search ad revenue was $60.4 billion, continuing to provide stable cash flow.
Google’s greatest advantage lies in its vast AI-era data assets—products like Search, YouTube, Maps, and Android have accumulated interaction data from over 2 billion users. Management noted that computing power remains limited in the short term; with more compute, cloud revenue would be even higher. The company has raised its 2026 capital expenditure guidance to $180-$190 billion.
Key market questions include: Will AI search impact traditional search ad revenue? How quickly can Gemini be commercialized? Can Google Cloud continue as the engine for AI growth? The Q2 2026 earnings report will be a critical checkpoint for these issues.
Amazon: Can AWS Redefine the AI Cloud Market?
Amazon (AMZN)
AWS remains the world’s largest cloud platform and the centerpiece of Amazon’s AI strategy.
In Q1 2026, Amazon’s revenue was $181.5 billion, up 17% year-over-year. AWS revenue reached $37.6 billion, up 28%, marking the fastest growth in 15 quarters. Amazon’s annualized AI revenue exceeded $15 billion, accounting for about 10% of AWS’s total. Citi estimates that AI revenue will account for about 58% of AWS’s incremental revenue in 2026 and about 72% in 2027. Analysts’ average price target is $314.23. In Q1 2026, Amazon’s earnings per share reached $2.78, up 74.84% year-over-year, with net profit at $30.255 billion, up 76.65%.
Amazon’s AI opportunities mainly lie in three areas: infrastructure demand for enterprise AI deployment, AI compute leasing services, and AI platform offerings like Amazon Bedrock. Goldman Sachs projects AWS revenue growth of about 33% year-over-year in 2026 and about 35% in 2027.
However, AI revenue’s share of AWS (about 10%) still lags behind Azure’s more than 20%. Whether Amazon can close the gap with Microsoft in AI cloud services will directly impact its valuation upside.
AMD and Broadcom: AI Chip Competition Enters the Second Phase
AMD (AMD)
AMD’s MI series AI chips are attempting to challenge NVIDIA’s dominance in the data center market. Analysts’ average price target is $544.90. AMD plans to host the Advancing AI 2026 event in San Francisco next week.
AMD’s key challenge: Can its AI chips achieve large-scale deployment, given NVIDIA’s strong ecosystem barriers?
Broadcom (AVGO)
Broadcom focuses on AI network chips and custom ASICs, serving the tailored needs of hyperscale cloud customers. As cloud providers accelerate in-house chip development, Broadcom’s strategic value as a custom chip design service provider is rising.
The future AI chip market may transition from "NVIDIA’s sole dominance" to a diversified era of "GPUs + ASICs + specialized chips." For investors, this means AI chip investment opportunities are expanding beyond NVIDIA and spreading across the entire semiconductor value chain.
The Biggest Risk for AI Stocks: Capital Expenditure Growth Outpacing Revenue Growth
Although AI revenue has surpassed cost lines, the industry is far from mature profitability. Investors should pay close attention to three major risks:
First, the risk of a capital expenditure bubble. In 2026, the combined capital expenditure guidance of Microsoft, Google, Amazon, Meta, and Oracle—the five hyperscale cloud providers—exceeds $750 billion. If data center investment continues to outpace AI revenue growth, it could lead to declining profit margins and even force companies to cut capital spending. Bloomberg’s consensus analyst forecasts suggest this figure could rise further to nearly $900 billion next year.
Second, downward pressure on AI model pricing. As the number of open-source models increases and inference costs continue to fall, profit margins for AI services may gradually shrink. The Scaling Law is hitting a wall: simply hoarding compute and increasing data volume is yielding diminishing returns in model performance.
Third, competition is shifting from technology to business efficiency. The future winners may not be the companies with the largest models, but those that can lower costs, improve efficiency, and generate sustained cash flow. The logic of secondary market valuations has already validated this trend: "The market no longer asks how big your model is, but whether your technology can deliver real-world results."
Key Metrics for AI Stock Investment in 2026
| Metric | Significance |
|---|---|
| AI Revenue Growth | Measures the true strength of commercial demand |
| AI Capital Expenditure | Gauges the sustainability of long-term competitiveness |
| Cloud Business Growth | Core channel for AI commercialization |
| Data Center Profit Margin | Assesses return efficiency on capital investment |
| Enterprise AI Adoption Rate | Determines the ceiling for long-term market size |
Conclusion
The AI industry is entering a brand-new phase of development.
In the past, investors bet on AI’s future potential. Now, the market is searching for the real cash flow that AI can generate.
NVIDIA represents the foundation of AI infrastructure. Microsoft and Google showcase the commercialization capabilities of AI platforms. Amazon demonstrates the scale effect of the cloud computing gateway. AMD and Broadcom embody the diversified competition in next-generation chips.
In the coming years, the core competition among AI stocks will no longer be about who owns the largest model parameters, but who can turn artificial intelligence into sustained commercial revenue growth. The Q2 2026 earnings season is evolving into a "real performance verification." For investors, shifting from "betting on stories" to "verifying data" will be the most rational investment strategy for this phase.
FAQ
Q1: What does it mean for AI revenue to surpass infrastructure costs?
It means the AI industry has crossed its first threshold for self-sustaining growth. In Q1 2026, global AI quarterly revenue was about $25 billion, surpassing the infrastructure depreciation cost of $21 billion for the first time. Cash flow generated by AI operations now covers the accounting depreciation costs for servers, GPUs, and data centers, marking the shift from pure investment to self-verification.
Q2: Is NVIDIA’s valuation already too high?
For fiscal year 2026, NVIDIA’s revenue was $216 billion, data center revenue $194 billion, and operating cash flow $103 billion. With a market cap of about $4.91 trillion, its valuation is indeed at a high level. The key issue is whether GPU demand can continue to translate into high profit growth. Competition in the AI chip market is intensifying, with AMD, cloud providers developing their own chips, and custom ASICs all vying for market share. This is the main risk NVIDIA faces.
Q3: Who has the edge in AI commercialization, Microsoft or Google?
Their approaches differ. Microsoft’s strength lies in deep integration between Azure and OpenAI, with annualized AI revenue reaching $37 billion. Google’s advantage is its massive data assets accumulated from the world’s largest search ecosystem, with Google Cloud revenue up 63% year-over-year to $20.03 billion. Microsoft has a deeper enterprise presence, while Google excels in data and search. Both have their strengths in AI commercialization, and it’s difficult to declare an absolute winner in the short term.
Q4: What is the biggest risk for AI stocks?
The biggest risk is capital expenditure growth consistently outpacing revenue growth. In 2026, the combined capital expenditure of the five major cloud providers exceeds $750 billion. If AI revenue growth slows while capital spending keeps expanding, profit margins will decline and valuations will come under pressure. Additionally, falling AI model prices and intensified open-source competition are significant risk factors.
Q5: What is the investment logic for AI stocks in the second half of 2026?
The core logic is shifting from "betting on AI’s future" to "verifying AI’s cash flow." Investors should focus on the actual growth rate of each company’s AI revenue, trends in cloud business profit margins, and the return efficiency on capital expenditures. The Q2 2026 earnings season will be a crucial window to assess these key indicators. Companies with clear commercialization paths and sustained profitability are most likely to see their valuations re-rated in the next phase.
