Global spending on AI infrastructure is accelerating at an unprecedented pace, driven by major tech firms and rising investor interest, yet questions about financial returns and market risks loom large amid a rapidly evolving landscape.
The ongoing debate over whether artificial intelligence (AI) represents a speculative bubble or a lasting technological revolution continues to engage investors and industry watchers alike. However, what remains clear is that spending on AI infrastructure and innovation is accelerating at an unprecedented pace, driven largely by major technology companies and hyperscalers. These firms are ramping up their capital expenditures aggressively, underscoring the strategic bet they are placing on AI as a core driver of future growth.
According to a recent analysis, the largest hyperscalers are expected to increase AI-related capital expenditures by 50% to reach around $450 billion in 2024. This uptick reflects the investments necessary to support next-generation AI workloads, particularly in data centre infrastructure optimized for AI training. A report from Dell’Oro Group further highlights the surge, noting a 51% growth in global data centre capital expenditure to $455 billion this year, with the top 10 hyperscalers accounting for more than half of that spend. Key technology advancements, such as NVIDIA’s Hopper and Blackwell architectures, alongside custom accelerators from Google, Amazon, and Microsoft, are major contributors to this investment boom.
Financial institutions project even more expansive growth over the coming years. Citigroup, for example, recently revised its forecast for AI infrastructure spending to surpass $2.8 trillion by 2029, up from an earlier $2.3 trillion estimate. This increase is fueled not only by hyperscalers but also escalating enterprise demand as real-world use cases validate AI’s value across sectors, including companies like Eli Lilly and Hitachi. Citigroup estimates that AI computing will need an additional 55 gigawatts of power by 2030, translating into vast capital needs.
Globally, this expansive spending is not limited to U.S. tech giants. Chinese tech powerhouse ByteDance plans to invest over $20 billion in AI-related infrastructure and applications in 2025, aiming to maintain its leadership in China’s AI space and bolster international capabilities despite regulatory uncertainties around its flagship app TikTok. This move underscores how AI is a strategic priority across multiple markets amid intensifying global competition.
Yet, amid soaring investment, significant questions about returns loom large. A recent J.P. Morgan report warns that the AI industry must generate around $650 billion in annual revenues by 2030 merely to achieve a 10% return on capital investment, a formidable challenge equating to substantial ongoing consumer payments if the revenue does not materialize. The report points to risks including consumer hesitation, potential overcapacity in AI infrastructure, and parallels to the telecom sector’s fiber buildout, which struggled with underutilized assets. It highlights that while AI may produce market titans, it could also leave substantial market value at risk if expectations are not met.
Within this rapidly evolving context, investors seeking exposure to AI and technology innovation face the challenge of selecting potential winners across a sprawling landscape dominated by a handful of mega-cap companies collectively dubbed the “Magnificent 7” or even “Magnificent 70,” given each firm’s ownership of multiple large businesses. For instance, Alphabet is much more than Google search, it encompasses YouTube, Google Ads, Chrome, cloud services, Waymo, Fitbit, Nest, and many others, each potentially a sizeable business in its own right.
For those looking to participate in the AI upside via a diversified approach, exchange-traded funds (ETFs) offer a convenient vehicle without the risk of single-stock exposure. On the Johannesburg Stock Exchange (JSE), several tech ETFs provide this broad market access. Notably, the ETF5IT from 1nvest offers passive exposure to the tech stocks embedded within the S&P 500, effectively representing the broad “Magnificent 70” tech conglomerates. Other more actively managed options on the JSE include Sygnia’s SYG4IR and SYFANG, which target stocks impacted by the Fourth Industrial Revolution and the core “FANG” companies respectively, as well as EasyETFs’ EASYAI focusing on AI innovation leaders.
According to a leading market commentator, holding a passive broad ETF like ETF5IT can provide balanced exposure given the presence of large AI-capable names, complementing diversified global holdings. Such approaches help mitigate concentrated risks while capturing the broad structural AI growth theme across industries and geographies.
The AI investment landscape also reflects a broader shift in the venture capital and startup ecosystem. Funding for AI and cloud companies in the U.S., Europe, and Israel is projected to hit $79.2 billion in 2024, an increase of 27% from the previous year, with about 40% of this directed at generative AI technologies. The U.S. dominates this trend, receiving 80% of generative AI investments, driving rapid scaling and commercialisation, while Europe and Israel capture the remainder, focusing on innovative startups seeking to carve out niches within the global AI wave.
Alphabet’s recent quarterly results exemplify the current complexities of the AI investment cycle. Despite strong revenue growth driven by AI-enhanced products and a robust advertising market, the company boosted its annual capital expenditure plan to approximately $85 billion. This move underlines the costly and competitive nature of infrastructure buildout associated with AI leadership, reflecting a broader industry pattern marked by increased borrowing to finance these outlays, even as investors question the pace and scale of AI monetisation.
In sum, AI infrastructure spending and investment continue to accelerate at a breathtaking scale globally, driven by hyperscalers, tech giants, and emerging players alike. While this momentum confirms AI’s transformative potential, it also raises fundamental questions about the timing and magnitude of financial returns, market valuation risks, and the selection of future winners. For investors, diversified access through tech and AI-focused ETFs remains an attractive and pragmatic way to engage with this profound technological shift amidst the evolving dynamics and uncertainties of the AI market.
- https://justonelap.com/jse-listed-ai-etfs/ – Please view link – unable to able to access data
- https://www.reuters.com/world/china/citigroup-forecasts-big-techs-ai-spending-cross-28-trillion-by-2029-2025-09-30/ – Citigroup has revised its forecast for AI-related infrastructure investments by major tech companies, expecting spending to surpass $2.8 trillion by 2029, up from a previous estimate of $2.3 trillion. The increase is attributed to early, aggressive investments by hyperscalers—large-scale data center operators such as Microsoft, Amazon, and Alphabet—as well as rising enterprise demand for AI capabilities. By 2026, AI capital expenditures are projected to reach $490 billion. This surge in spending follows the explosive growth of AI technologies after the launch of ChatGPT in late 2022, despite challenges like China’s development of cheaper alternatives and concerns about U.S. trade tariffs under President Donald Trump. Citigroup estimates that the global need for AI computing will require 55 gigawatts of additional power by 2030, costing around $2.8 trillion, with $1.4 trillion to be spent in the U.S. alone. Notably, tech firms are increasingly turning to borrowing rather than relying solely on profits to finance these vast infrastructure projects. This trend is impacting their free cash flows and raising questions among investors about sustainable funding methods. Citigroup also highlights that enterprises such as Eli Lilly, Hitachi, and Wolters Kluwer are already demonstrating real-world AI use, validating the technology’s value.
- https://www.reuters.com/technology/bytedance-plans-20-billion-capex-2025-mostly-ai-sources-say-2025-01-23/ – ByteDance, the Chinese owner of TikTok, plans to invest over 150 billion yuan ($20.64 billion) in capital expenditure in 2025, with much of the focus on artificial intelligence. Approximately half of this investment will be spent abroad on AI-related infrastructure such as data centers and networking equipment. Notable beneficiaries of this spending include Huawei Technologies, Cambricon Technologies, and U.S. supplier Nvidia. ByteDance aims to use this expenditure to maintain its AI leadership in China and bolster its AI capabilities internationally, especially as it faces uncertainties regarding TikTok’s future in the U.S. Recently, ByteDance has expanded its AI offerings extensively, with more than 15 standalone AI applications. The company operates several AI apps, including Doubao, a chatbot with 75 million monthly active users, and has introduced international versions of its key AI applications. The planned spending is substantial but still modest compared to U.S. tech giants like Alphabet and Microsoft.
- https://www.reuters.com/technology/artificial-intelligence/ai-cloud-funding-us-europe-israel-hit-79-bln-2024-accel-says-2024-10-16/ – In 2024, funding for artificial intelligence (AI) and cloud companies in the U.S., Europe, and Israel is expected to reach $79.2 billion, a 27% increase from $62.5 billion in 2023, according to a report from venture capital firm Accel. Investments in generative AI companies account for about 40% of this total. Philippe Botteri of Accel highlighted the significant impact of AI, surpassing previous advancements in broadband, mobile, or cloud technology. The U.S. is the primary recipient of AI investments, attracting 80% of the $56 billion invested in generative AI over the past two years, with major contributions from companies like OpenAI, xAI, and Anthropic. Europe and Israel received the remaining 20%, with notable funding for companies like Mistral, Aleph Alpha, and DeepL. Despite the strong performance in AI, the outlook for the broader software sector shows a shift towards profitability over high growth.
- https://www.prnewswire.com/news-releases/data-center-capex-surged-51-percent-to-455-billion-in-2024-according-to-delloro-group-302405209.html – According to a recently published report by Dell’Oro Group, the trusted source for market information about the telecommunications, security, networks, and data center industries, global data center capex surged 51 percent to $455 billion in 2024. Accelerated servers optimized for AI training workloads deployed by the hyperscalers accounted for most of the growth. “The top 10 hyperscalers accounted for more than half of global data center capex in 2024, driven largely by heightened investments in AI infrastructure,” said Baron Fung, Sr. Research Director at Dell’Oro Group. “While NVIDIA’s Hopper architecture—followed by Blackwell systems later in the year—dominated spending, custom accelerators from Google, Amazon, and Microsoft also fueled growth. Additionally, Tier 2 cloud providers like xAI and CoreWeave significantly ramped up their capex, approaching hyperscaler levels due to increased GPU deployments.
- https://www.reuters.com/business/google-parent-alphabet-beats-quarterly-revenue-estimates-boosts-capex-plan-2025-07-23/ – Alphabet Inc., Google’s parent company, exceeded Wall Street expectations by reporting $96.43 billion in total revenue for Q2 2025, surpassing the analyst consensus of around $94 billion. This strong performance was driven by new AI features and a stable digital advertising market. In response, Alphabet announced an increase in its annual capital expenditure plan to approximately $85 billion, up from a previously pledged $75 billion. This move reflects the broader trend among Big Tech firms, which are collectively projected to invest over $320 billion in AI infrastructure amid global competition and investor concerns over the pace of AI monetization. Despite the positive earnings report, Alphabet’s stock saw a 1% drop in after-hours trading, although shares have risen over 18% since the last earnings report in April.
- https://www.tomshardware.com/tech-industry/artificial-intelligence/usd650-billion-in-annual-revenue-required-to-deliver-10-percent-return-on-ai-buildout-investment-j-p-morgan-claims-equivalent-to-usd35-payment-from-every-iphone-user-or-usd180-from-every-netflix-subscriber-in-perpetuity – A recent J.P. Morgan report reveals that the AI industry must generate approximately $650 billion annually by 2030 to attain just a 10% return on anticipated capital investments. This projection, shared by analyst Max Weinbach, translates to ongoing monthly payments of $34.72 from each of the 1.5 billion iPhone users or $180 yearly from more than 300 million Netflix subscribers. Despite the promising revenue growth from companies like OpenAI and Anthropic, the report highlights significant risks, including consumer hesitation and potential overcapacity in AI infrastructure. The report draws a parallel to the telecom fiber buildout, warning that revenue might not keep pace with investment, potentially leading to idle AI data centers. While it stops short of calling it an “AI bubble,” the document acknowledges warnings from tech leaders about slow returns and market volatility. If the expected AI returns fail to materialize, nearly $20 trillion in global market cap could be at risk. Still, J.P. Morgan emphasizes that the AI space may produce both major winners and significant losers, reflecting the high-stakes, high-capital nature of the industry.
Noah Fact Check Pro
The draft above was created using the information available at the time the story first
emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed
below. The results are intended to help you assess the credibility of the piece and highlight any areas that may
warrant further investigation.
Freshness check
Score:
8
Notes:
The narrative references recent reports from Dell’Oro Group and Citigroup, dated March 2025 and September 2025 respectively, indicating a high freshness score. The article was published on 13 November 2025, suggesting timely reporting. However, the content appears to be a repurposed analysis of existing reports, with no new data or exclusive insights, which may indicate recycled content. Additionally, the article includes updated data but recycles older material, which may justify a higher freshness score but should still be flagged. ([prnewswire.com](https://www.prnewswire.com/news-releases/data-center-capex-surged-51-percent-to-455-billion-in-2024-according-to-delloro-group-302405209.html?utm_source=openai))
Quotes check
Score:
7
Notes:
The article includes direct quotes from reports by Dell’Oro Group and Citigroup. These quotes are consistent with the original reports, suggesting they are reused content. No new or exclusive quotes are provided, which may indicate a lack of originality. ([prnewswire.com](https://www.prnewswire.com/news-releases/data-center-capex-surged-51-percent-to-455-billion-in-2024-according-to-delloro-group-302405209.html?utm_source=openai))
Source reliability
Score:
6
Notes:
The narrative originates from Just One Lap, a reputable South African financial education platform. However, the article primarily repurposes information from Dell’Oro Group and Citigroup reports, which are reputable sources. The reliance on these secondary sources without additional verification or original reporting raises concerns about the reliability of the information presented. ([prnewswire.com](https://www.prnewswire.com/news-releases/data-center-capex-surged-51-percent-to-455-billion-in-2024-according-to-delloro-group-302405209.html?utm_source=openai))
Plausability check
Score:
7
Notes:
The claims regarding AI infrastructure spending and the availability of AI-focused ETFs on the Johannesburg Stock Exchange (JSE) are plausible and align with known industry trends. However, the lack of new insights or original reporting, combined with the recycling of older material, suggests that the narrative may not provide substantial new value to readers. Additionally, the article lacks specific factual anchors, such as names, institutions, or dates, which reduces its credibility. ([prnewswire.com](https://www.prnewswire.com/news-releases/data-center-capex-surged-51-percent-to-455-billion-in-2024-according-to-delloro-group-302405209.html?utm_source=openai))
Overall assessment
Verdict (FAIL, OPEN, PASS): FAIL
Confidence (LOW, MEDIUM, HIGH): MEDIUM
Summary:
The narrative primarily repurposes existing reports from Dell’Oro Group and Citigroup without providing new insights or original reporting. The reliance on secondary sources and the lack of exclusive content raise concerns about the freshness and originality of the information presented. While the claims are plausible and align with known industry trends, the overall lack of new value and the recycling of older material suggest that the narrative does not meet the standards for a PASS assessment.

