Picture this: a fund manager in Milan or Mumbai just watched an AI-linked stock swing 7% in a single earnings call, and now has to decide where to park the next billion dollars. So who are the AI winners likely to emerge from this next phase of the boom? According to Reuters, big institutional investors are shifting away from panicking over hyperscaler spending and are now actively hunting for the companies that will actually cash in on the artificial intelligence buildout, not just the ones writing the biggest cheques for data centers.
That shift , from “how much are they spending?” to “who will actually profit?” , is the central story of AI investing in the second half of 2026. If you’re a student, a fresher, or a young professional in Odisha trying to make sense of the AI economy, understanding this shift matters, because it tells you where jobs, funding, and opportunity are likely to concentrate next.
What Sparked the “AI Capex Angst” in the First Place?
Capex (capital expenditure) is the money a company spends on long-term physical assets , think data centers, chips, and power infrastructure , rather than day-to-day operating costs. For AI, capex mostly means building and equipping the giant server farms that train and run large language models.
Through mid-2026, that spending exploded. Alphabet’s second-quarter capex came in at $44.9 billion, and the company raised its full-year guidance to a range of $195 billion to $205 billion, up from an earlier $180–190 billion, with executives flagging a “significant” further increase expected in 2027, according to Yahoo Finance’s earnings coverage. Investors did not celebrate. Alphabet shares fell 7.13% on the news, wiping out roughly $293 billion in market value in a single session. Tesla’s announcement of a $25 billion capex commitment for 2026 , about three times its historical spending , triggered an even sharper 14.5% stock decline the same day.
That is what “capex angst” looks like: markets punishing companies for spending aggressively on AI, even when the underlying business is growing, because investors worry the payoff is too far away or too uncertain.
Why Is Capex Angst Fading Now?
Is the AI spending backlash over? Not entirely, but the mood has clearly cooled. Reuters reports that the rally in AI-linked stocks through the latest earnings season has shifted its focus , investors are no longer simply reacting to capex headlines with knee-jerk sell-offs. Instead, they are digging into which companies show a credible path to monetizing that spending.
Part of this comes down to evidence. Big Tech firms have started showing that AI is generating real revenue, not just cost. Analysts covering the sector have pointed out that Alphabet, Microsoft, and Amazon are each reporting rising cloud and AI-services revenue alongside their capex increases , which is why some investors are willing to look past the sticker shock of the spending itself and focus on the return on that spending instead.
There’s also a psychological shift at play. Earlier in the AI buildout, roughly between 2023 and mid-2025, markets tended to reward almost any company that announced large AI investments, treating spending itself as a signal of ambition and competitive strength. Michael Reynolds, vice president of investment strategy at Glenmede, described this as a reversal of an older pattern in which “asset-light” companies , those that didn’t need heavy capital investment , were the ones Wall Street preferred, according to Reuters coverage of the sector. Now, he noted, “this year is very much an outlier where those companies that are deploying capex at large scale are outperforming to a significant degree,” while cautioning that investors still need confidence that demand justifies the spend.
That distinction , spending as a signal of ambition versus spending as a proven path to revenue , is exactly what separates the current, more selective phase of AI investing from the earlier, more euphoric one. Investors aren’t abandoning the AI trade; they’re becoming pickier about which AI winners actually deserve their capital.
Where Are Big Investors Looking for Tomorrow’s AI Winners?
The hunt for the next wave of AI winners has moved well beyond the handful of household names that dominated headlines in 2023 and 2024. Investors are now spreading bets across several distinct categories.
Power and Infrastructure Providers
Why are energy companies suddenly “AI stocks”? Because data centers cannot run without electricity, and demand for power is now a bigger bottleneck than demand for chips in many markets. In a January 2026 investor survey covered by Reuters, asset manager BlackRock found that more than half of its EMEA clients backed providers of power needed by data centres as their preferred way to invest in the AI theme, while 37% named infrastructure as their top choice. Only about a fifth of respondents said the largest U.S. tech groups themselves were still the most compelling AI investment.
Inference and “Application Layer” Companies
Definition + Expansion , Inference: Inference is the stage where a trained AI model is actually put to use, generating answers, recommendations, or content for real users, as opposed to “training,” which is the earlier, more compute-heavy process of teaching the model in the first place. As AI shifts from a training-heavy phase to a usage-heavy phase, companies that make inference faster, cheaper, or more personalized , memory systems, specialized chips, optimization software , are drawing fresh investor attention, according to Axios reporting on how the AI trade is evolving in 2026.
Non-Tech Companies Using AI to Cut Costs
Some strategists argue the biggest AI winners won’t be tech companies at all. Bank of America’s global strategist Haim Israel has argued that framing AI purely around consumer chatbots like ChatGPT or Gemini misses the bigger picture , much like judging the entire internet revolution by Spotify alone. Old-economy companies that use AI internally to cut costs and improve efficiency are increasingly part of the AI investment story, per Axios.
Broader Geographic and Sector Diversification
Institutional investors are also spreading their AI bets geographically and across sectors rather than concentrating purely on a handful of U.S. megacaps, a trend that has been building since emerging-market fund managers first started looking for AI supply-chain plays beyond Nvidia.
Comparing Where Investor Money Is Flowing
| Investment Category | What It Covers | Why Investors Like It Now | Key Risk |
| Power & Data Center Infrastructure | Utilities, grid operators, cooling and power-supply makers | Bottleneck on AI growth is electricity, not just chips | Regulatory delays, long build timelines |
| Inference & Memory Tech | Chips and software that make running (not training) AI models cheaper | ROI is now the priority over raw scale | Fast-moving, hard to pick durable winners |
| Hyperscalers (Big Tech) | Alphabet, Microsoft, Amazon, Meta capex and cloud revenue | Still the biggest, most liquid AI exposure | Capex guidance shocks hit share prices hard |
| AI-Adopting “Old Economy” Firms | Retail, logistics, manufacturing using AI to cut costs | Real, provable efficiency gains | AI benefit is harder to isolate from other factors |
| Emerging Markets AI Supply Chain | Component makers, semiconductor suppliers outside the US | Better relative value, wider opportunity set | Currency and geopolitical exposure |
How Are Analysts Actually Evaluating a Potential AI Winner?
What separates a genuine AI winner from a company just riding the hype? Increasingly, analysts and fund managers are applying a more disciplined checklist rather than simply buying whatever is labeled “AI.” A few criteria show up repeatedly across recent coverage from Reuters, Axios, and Wall Street research desks:
- Demonstrated revenue tied to AI, not just spending announcements , for example, cloud-computing growth that can be directly linked to AI workloads.
- Return on invested capital (ROI) clarity , whether management can show, quarter over quarter, that new data-center capacity is translating into paying customers rather than idle compute.
- Balance sheet discipline , how much of the AI buildout is funded through free cash flow versus new debt or secondary equity offerings, since heavier borrowing raises risk if returns are delayed.
- Exposure to bottlenecks, not just growth , companies tied to power generation, grid capacity, or advanced cooling are increasingly viewed as safer, more foundational bets than pure software plays.
- Diversified customer base , reliance on just one or two large AI labs or hyperscalers as customers is seen as a concentration risk, especially for smaller suppliers.
This more rigorous framework is precisely why the list of likely AI winners now spans far beyond the original handful of chipmakers and cloud giants that dominated headlines in the early years of the boom.
The Hyperscaler Capex Race: By the Numbers
Understanding the scale of hyperscaler spending helps explain why investors are being more selective about AI winners rather than simply buying every AI-linked stock. Some key figures from 2026 earnings season, per Reuters and industry trackers:
- S&P 500 capital expenditure plans have ballooned to roughly $1.2 trillion this year , the highest level since research firm Trivariate began tracking the data in 1999, with the top nine companies accounting for nearly 30% of the total, according to Reuters’ Johann M. Cherian.
- Total shareholder returns (dividends plus buybacks) across the S&P 500 also hit a record $1.65 trillion for the 12 months ending in June 2026, per S&P Global data cited by Reuters , showing that heavy AI spending hasn’t fully crowded out payouts to shareholders.
- Alphabet’s full-year 2026 capex guidance now stands at $195–205 billion, up from an earlier $180–190 billion range.
- Amazon’s capex hit roughly $41 billion in a recent quarter, up 70% year-over-year, as it accelerated data-center expansion, according to AlphaSense’s tracking of Big Tech AI spending.
- Tesla committed $25 billion in 2026 capex , about three times its historical annual spending , largely tied to Optimus and robotaxi production.
- Credit spreads for Google, Amazon, and Meta have been widening in the bond market, meaning fixed-income investors are demanding more compensation to lend to these companies as AI-related debt and spending climb, CNBC reported in July 2026.
Nvidia’s New Playbook: From Chipmaker to Ecosystem Financier
Is Nvidia still the obvious AI winner it used to be? Largely yes, but its role is expanding beyond simply selling chips. On August 10, 2026, Nvidia announced partnerships with major Wall Street firms aimed at mobilizing more than $500 billion of third-party capital to help fund the broader buildout of AI infrastructure , effectively positioning the company as a financier and ecosystem-builder for the industry, not just a hardware vendor. Nvidia CEO Jensen Huang has separately said he expects AI data-center spending to continue for seven to eight years, signaling that the current investment cycle is still in its early-to-middle stages rather than near its end.
This shift matters for anyone trying to spot future AI winners: the companies that control access to capital and infrastructure financing, not just those that sell the fastest chip, may end up shaping who gets to build , and who gets left behind , in the next phase of AI expansion.
Risks Investors Are Still Watching
Even as capex angst fades, it hasn’t disappeared. A few risk factors remain squarely on investors’ radar:
- Widening bond spreads for the biggest AI spenders suggest lenders are pricing in more risk around debt-funded data-center buildouts, per CNBC’s July 2026 coverage.
- Rising secondary equity offerings , Wall Street Horizon’s corporate event data reportedly showed Q2 2026 featured the most secondary equity offerings in five years, a signal some analysts read as companies scrambling for capital to fund AI ambitions.
- Depreciation drag: today’s massive capex eventually becomes tomorrow’s depreciation expense, spread across future earnings , a headwind unless AI-driven revenue growth keeps pace.
- Persistent bubble talk: even optimistic analysts, including strategists at Morgan Stanley Wealth Management, have flagged that markets may eventually start questioning whether AI’s promise is “fully priced” into current valuations.
What This Means for AI Talent and Businesses in India
For students and young professionals in Odisha and across India, this rebalancing among AI winners has a practical takeaway: the opportunity is broadening beyond just “learn to prompt ChatGPT.” As investment spreads into power infrastructure, inference optimization, applied AI in traditional industries, and emerging-market supply chains, the skill sets in demand are diversifying too , from AI engineering and MLOps to energy-sector data roles and applied AI for manufacturing, retail, and logistics.
This is exactly the shift Kalinga.ai’s training programs are built around: helping learners move from foundational AI literacy toward the kind of applied, agentic AI engineering skills that companies across sectors , not just Big Tech , are now hiring for.
Key Takeaways
- Big investors are moving from fearing AI capex to hunting for AI winners that can prove real returns on that spending.
- Power, infrastructure, and inference-layer companies are drawing fresh attention alongside traditional hyperscalers.
- S&P 500 capex has hit roughly $1.2 trillion, the highest level since 1999, even as shareholder payouts also hit records.
- Nvidia’s shift toward financing AI infrastructure (over $500 billion mobilized with Wall Street partners) shows the industry’s next phase is as much about capital access as chip performance.
- Risks , widening bond spreads, rising equity issuance, and depreciation drag , mean the search for AI winners is far from risk-free.
FAQ: Understanding the Search for the Next AI Winners
What does “AI capex angst” mean? It refers to investor anxiety over the enormous capital expenditure , spending on data centers, chips, and power infrastructure , that major tech companies have committed to AI, and uncertainty over whether that spending will generate adequate returns.
Why did Alphabet’s stock fall after raising its AI spending guidance? Alphabet’s shares dropped 7.13%, erasing about $293 billion in market value, after the company raised its 2026 capex guidance to $195–205 billion and flagged a further significant increase for 2027, which investors read as a signal of rising costs without immediate clarity on returns, according to Yahoo Finance.
Which sectors are investors now favoring for AI exposure? Beyond the major hyperscalers, investors surveyed by BlackRock in January 2026 favored power and data-center infrastructure providers, with over half naming this as a top choice, alongside broader infrastructure plays and inference-layer technology, per Reuters and Axios reporting.
Is Nvidia still considered a top AI winner in 2026? Yes , Nvidia remains central to the AI buildout, and has expanded its role by mobilizing over $500 billion in third-party capital with Wall Street partners as of August 2026, alongside CEO Jensen Huang’s outlook that AI data-center spending will continue for another seven to eight years.
What risks could derail the current AI investment rally? Widening bond spreads for major AI spenders, a rise in secondary equity offerings to fund capex, and the eventual depreciation drag from today’s infrastructure spending are all risks investors are actively monitoring, per CNBC and Wall Street Horizon data.
How does this trend affect AI careers in India? As investment spreads from pure infrastructure into inference, applied AI, and non-tech adopters of AI, demand for talent is broadening too , creating openings in applied AI engineering, data roles, and AI-driven operations across industries beyond Big Tech.
Curious how these shifts translate into real career and business opportunities closer to home? Explore Kalinga.ai’s latest AI industry coverage and our hands-on LLM Engineering and agentic AI workshops designed for students and professionals across Odisha.