Enjoying the read? Subscribe for free and stay informed.

Join over 300,000 readers who trust us to stay ahead!

News/Commentary

LOOKING BACK | Is the AI Capital Gravy Train Running Out of Steam?

By DWN Staff · Nov 19, 2026 Views: 3,803

LOOKING BACK | Is the AI Capital Gravy Train Running Out of Steam?
Did we really reach peak AI this week? Probably not, but we likely wouldn't know until well long after a peak has been achieved.

The days are short, the nights are long, it's cold outside, and it's time to party. It's no coincidence that cultures across the temperate zones on our planet tend to bunch their holiday celebrations into the short, cold days, we all need good times to see us through harder days.  

But what if, just as the days are cooling off, the party is ending in artificial intelligence capital funding?  

Artificial intelligence has become one of the greatest magnets for investment capital in financial history, drawing money from venture capital firms, private equity funds, sovereign wealth funds, pension funds, technology companies, banks and individual investors. Yet the industry's extraordinary ability to attract financing is increasingly accompanied by an equally extraordinary question: How much money can artificial intelligence consume before investors begin demanding more convincing returns? Recent developments involving OpenAI and Anthropic illustrate the apparent contradiction at the heart of the AI investment boom. The industry's leading companies are pursuing financing on a scale that would have seemed unimaginable only a few years ago, even as investors become more concerned about profitability, debt, infrastructure spending and valuations. For financial services professionals, the debate is no longer simply whether AI represents a transformative technology. It is whether the capital markets can finance that transformation profitably—and whether investors entering the market today can still capture the returns enjoyed by its earliest backers.  

The latest fundraising developments at OpenAI suggest that even the most prominent AI developers cannot assume investors will provide unlimited financing on whatever terms they demand. Reports in late October indicated that OpenAI had reduced the target for a proposed funding round from $30 billion to approximately $20 billion, after reportedly securing about $11 billion in commitments. That development, if interpreted in isolation, might suggest that enthusiasm for AI investment is weakening. However, a $20 billion financing would still constitute an extraordinary fundraising achievement for almost any private company. The more significant issue is the relationship between the amount of capital OpenAI needs, the valuation it seeks and the economic returns its investors can reasonably anticipate.  

Anthropic represents the other side of the equation: the possibility that the enormous private valuations assigned to leading AI developers can be validated by public investors. Its reported November public offering, described in the supplied November 18 and 19 Reuters coverage as a blockbuster transaction, makes the company an important test of whether public markets are prepared to absorb the capital requirements and valuations of 

For financial artificial intelligence, the significance of these developments extends far beyond two companies. OpenAI and Anthropic are suppliers of foundational technology increasingly incorporated into banking, insurance, asset management, wealth management, financial research and enterprise software. Their ability to finance increasingly sophisticated models will influence the price, availability, capabilities and competitive structure of AI services used throughout the financial industry. If capital becomes more expensive, financial institutions may face higher AI operating costs, slower model improvements or greater dependence on a smaller number of well-capitalized technology providers. Conversely, successful fundraising and public-market participation could support another generation of increasingly powerful financial AI applications. 

An Unprecedented Concentration of Investment Capital 

The numbers underlying the AI fundraising boom are difficult to reconcile with the notion that investors have abandoned the industry. According to the October 2026 PitchBook–National Venture Capital Association Venture Monitor, U.S. venture capital deal value reached $515.8 billion during the first three quarters of 2026, approximately 44% above the previous full-year record established in 2021. AI companies accounted for an extraordinary 82.7% of that activity. That concentration indicates that AI is no longer merely one promising sector among many competing for venture investment. It has become the dominant destination for American venture dollars 

However, interpreting these figures requires distinguishing between money invested in companies, money raised by investment funds and money spent on physical infrastructure. Venture capital deal value measures investments made into startups. Venture fundraising measures commitments investors make to venture funds, which may subsequently be deployed over several years. Corporate capital expenditures represent spending on equipment, facilities and other assets. Private credit and bond issuance represent borrowed capital rather than equity investment. Adding these figures together without accounting for overlapping transactions can substantially exaggerate the total amount of financing available to AI. 

This distinction is particularly important because the AI economy increasingly depends on multiple financing channels. A frontier model developer might raise equity from venture investors, purchase computing services from a hyperscaler, and rely on that hyperscaler to finance data centers through corporate borrowing or private infrastructure funds. The same economic activity can consequently appear in several fundraising and investment categories. Investors evaluating the industry's financial health should focus not only on headline commitments but also on where the money originates, how it is ultimately spent and whether the resulting revenues can support the financing structure. 

The concentration extends to the identity of investors. A July PitchBook analysis found that corporate venture capital accounted for 87.9% of U.S. AI venture deal value during the period examined in 2026. The report identified Nvidia, Microsoft, Amazon, Google and other strategic investors as major forces in AI financing. These companies are not simply searching for attractive financial returns. They also have commercial incentives to support developers that purchase their chips, use their cloud infrastructure or incorporate their technologies. 

This introduces an important complication for conventional venture capital analysis. Some investors may be willing to accept lower direct returns from their equity investments because they anticipate additional benefits from hardware sales, cloud subscriptions, enterprise relationships or technological partnerships. An investment that looks questionable when evaluated independently might make more sense as part of a larger corporate strategy. Conversely, these relationships can obscure the underlying profitability of the AI ecosystem and make it difficult to determine whether demand is sustainable without continuing financial support from strategic partners. 

Is AI Fundraising Accelerating or Slowing Down? 

The answer depends on what is being measured. By aggregate dollars invested, AI fundraising has expanded enormously. By the number of companies receiving financing, the breadth of participation and the ease with which less-established startups can obtain capital, the picture is more complicated. 

The industry's trajectory increasingly resembles a market in which larger amounts of money are flowing toward fewer companies. Frontier model developers, major infrastructure providers and established enterprise AI businesses can attract extraordinary investments, while smaller startups face more demanding investors, longer fundraising cycles and greater pressure to demonstrate actual revenue. Consequently, a decline in the number of deals can occur simultaneously with an increase in total funding. 

This helps explain the conflicting conclusions circulating in investment forums, social media and technology commentary. Skeptics point to OpenAI's reduced fundraising target, declining valuations for some private technology companies, expensive computing infrastructure and limited evidence of profits sufficient to justify the industry's capital requirements. They interpret these developments as signs that investors are losing patience. 

Optimists emphasize the enormous investments still being completed, Anthropic's successful public offering, rising enterprise adoption and the possibility that AI will eventually transform most economic activities. From this perspective, investors are not withdrawing from AI. They are becoming more discriminating about which companies deserve their money. 

Both interpretations contain elements of truth. A reduction in speculative financing does not necessarily represent a contraction in the industry's total investment. Similarly, record fundraising does not establish that all AI investments are economically justified. 

The distinction becomes especially important when comparing quarterly activity. A single $50 billion or $100 billion transaction can radically change aggregate fundraising statistics. If such a transaction occurs in one quarter but not the next, the apparent decline may reflect the timing of a few unusually large deals rather than a meaningful deterioration in investor demand. 

The public-market reception for Anthropic provides a particularly revealing example. Reuters reported that the company raised approximately $81.3 billion in its November IPO, with an additional $26 billion private placement bringing the combined transaction to approximately $107.3 billion. Its shares were priced at $83 and closed their first trading day at $97.17, a 17% increase. That performance suggests considerable investor demand, although the stock retreated from a substantially higher opening price. 

The offering demonstrates that capital remains available for exceptional AI businesses, but it does not establish that comparable financing is available throughout the industry. Anthropic's ability to attract public investors at an enormous valuation says relatively little about the prospects of an early-stage company developing a narrow AI application with limited competitive advantages. 

Why the Skeptics Think the Money Is Running Out 

The bearish argument begins with a straightforward observation: Developing advanced artificial intelligence is extraordinarily expensive, and the industry's costs are increasing faster than many observers anticipated. 

Frontier AI developers require enormous computing resources to train and operate their models. Data centers need advanced semiconductors, networking equipment, cooling systems, reliable electricity, land, construction materials and specialized labor. Training increasingly sophisticated models can require investments that only a handful of organizations can afford. 

Operating those models creates additional expenses. Unlike conventional software, which can often serve additional customers at relatively low marginal cost, many AI applications incur meaningful computing costs every time a user submits a request. Although model efficiency is improving, growing usage and more computationally intensive reasoning can offset those savings. 

The result is a business model that may require sustained investment long before its ultimate profitability becomes clear. Anthropic, for example, entered public markets with substantial infrastructure commitments and continued losses, according to Associated Press reporting on its offering. The company reportedly faced approximately $900 billion in computing and infrastructure commitments over the coming years, illustrating the extraordinary scale of its ambitions and obligations. 

Critics also question whether AI customers will ultimately pay enough to justify these investments. Enterprises may enthusiastically experiment with AI without being willing to purchase expensive subscriptions indefinitely. Consumers accustomed to free or inexpensive AI services may resist higher prices. Competitive pressure among model developers can also make it difficult for providers to increase prices even as their capabilities improve. 

Another concern involves technological commoditization. If increasingly capable models become widely available from multiple providers, the economic value of basic intelligence services could decline. Open-source models, specialized smaller models and efficiency improvements may reduce the need for customers to purchase the most expensive frontier systems. 

Investors also face valuation risk. A company valued at $1 trillion must generate extraordinary future earnings to deliver attractive long-term returns. Even if its technology succeeds commercially, investors purchasing shares at an exceptionally high valuation may earn disappointing returns. 

Finally, higher interest rates and tighter financial conditions increase the cost of financing long-duration projects. AI infrastructure investments may not generate meaningful cash flows for years, making their valuations sensitive to changes in borrowing costs and expected returns. A business that appears attractive when capital is cheap can become much less compelling when investors demand higher yields. 

Why the Optimists Believe Fundraising Could Accelerate Again 

The bullish argument rests on the possibility that AI is still in the early stages of a technological transformation comparable to electrification, industrial automation or the commercialization of the internet. 

If AI becomes a foundational component of business operations, demand for computing infrastructure could continue expanding for many years. Financial institutions alone have opportunities to deploy AI across research, compliance, customer service, cybersecurity, fraud detection, underwriting, investment management and administrative workflows. 

Supporters also argue that the current emphasis on infrastructure is a prerequisite for future productivity gains. Railroads, telecommunications networks and cloud computing required enormous initial investments before their full economic benefits became apparent. The fact that AI infrastructure is expensive does not necessarily mean that it is being overbuilt. 

Anthropic's reported revenue trajectory offers evidence supporting this argument. Reuters reported that its annualized revenue increased from approximately $1 billion in early 2025 to $81 billion in 2026. Although annualized revenue is not the same as audited full-year revenue or profit, the reported growth suggests that commercial demand for frontier AI services can expand rapidly. 

There is also the prospect of new technological capabilities creating entirely new markets. More reliable AI agents could automate complex business processes, scientific research and software development. Robotics could extend AI into manufacturing, transportation, healthcare and logistics. Advances in model efficiency might simultaneously lower operating costs and expand the range of economically viable applications. 

Under this scenario, the industry's capital requirements could rise rather than decline. As AI becomes more useful, more businesses may demand access to it, encouraging further investment in computing infrastructure and specialized applications. 

Fundraising could also accelerate through the development of new financing structures. Rather than relying exclusively on venture capital, AI companies increasingly have access to corporate bonds, private credit, infrastructure funds, project financing, equipment-backed lending and public equity markets. The expansion of these channels could support investments far beyond what conventional venture capital funds can finance independently. 

However, greater access to capital does not eliminate economic risk. It redistributes that risk among shareholders, lenders, infrastructure investors and ultimately the customers whose payments must support the system. 

Is Fundraising Slowing Across the Entire Economy?  

AI's extraordinary fundraising performance contrasts sharply with the challenges confronting many other investment sectors. According to S&P Global Market Intelligence, global private equity fundraising declined 11% in 2025 to approximately $490.8 billion, marking a second consecutive annual decline. Venture capital fund fundraising also weakened, reflecting limited distributions, difficult exit conditions and institutional investors' reluctance to commit additional capital before receiving proceeds from existing investments.  

However, the broader picture improved somewhat during 2026. S&P Global reported that private equity firms raised approximately $306.8 billion during the first half of the year, putting the industry on a trajectory to reverse its earlier declines. The recovery was uneven, with established managers and firms capable of returning capital to investors enjoying a considerable advantage over newer competitors.  

This resembles the emerging pattern in AI. Investors have not necessarily stopped allocating capital to private markets. Instead, they are concentrating their commitments among managers and companies they perceive as offering the strongest prospects.  

The contrast raises a question about whether AI is consuming capital that would otherwise support biotechnology, healthcare, manufacturing, energy, financial technology and other industries.  

There is considerable evidence of concentration. The Organization for Economic Cooperation and Development reported that AI accounted for 61% of global venture capital investment in 2025, representing $258.7 billion of approximately $427.1 billion invested worldwide. AI's share had roughly doubled from 30% in 2022. The OECD also found that deals exceeding $100 million represented approximately 73% of AI venture investment value in 2025.  

This concentration creates genuine challenges for non-AI startups competing for investors' attention. A venture capitalist allocating additional capital to frontier AI companies may have less money available for a promising medical device manufacturer, financial technology startup or conventional software business.  

Nevertheless, describing AI as starving other sectors requires qualification. Investment capital is not a permanently fixed pool. Successful AI investments can create wealth that is subsequently reinvested elsewhere. AI can also stimulate investment in adjacent industries, including energy generation, semiconductor manufacturing, telecommunications, construction and cybersecurity.  

Furthermore, many companies classified as AI investments operate in other industries. An AI drug discovery company may attract funding that would otherwise be classified as biotechnology investment. An AI-enabled financial software provider may be counted as both fintech and artificial intelligence.  

The more important question is whether investors are neglecting economically attractive opportunities outside AI because they expect superior returns from the industry's most fashionable companies.  

The Financial Services Industry's Growing Exposure  

For financial institutions, AI fundraising presents both commercial opportunities and financial risks.  

Investment banks can earn substantial fees underwriting AI public offerings, arranging debt financing and advising on mergers and acquisitions. Private equity firms can acquire AI-related businesses or finance infrastructure development. Asset managers can create investment strategies targeting companies that benefit from AI adoption. Wealth managers can provide clients with exposure to the technology through public equities, diversified funds and selected private investments.  

However, financial institutions also face growing concentration risk. AI-related businesses increasingly influence major equity indexes, technology sector performance and the valuations of private portfolios. A substantial correction in AI valuations could consequently affect investors who have little direct exposure to frontier AI startups.  

Credit exposure deserves particular attention. Data center financing increasingly involves long-term leases, specialized equipment and expectations of sustained demand for computing capacity. If utilization disappoints, lenders may discover that collateral values are less stable than anticipated.  

Rapid technological obsolescence complicates the situation. Advanced processors can lose economic value as more efficient alternatives become available. Financial institutions evaluating equipment-backed loans must therefore distinguish between an asset's accounting depreciation schedule and its realistic resale value.  

Financial advisors should also examine how AI exposure overlaps across client portfolios. An investor might own shares in Microsoft, Amazon, Alphabet and Nvidia, hold an AI-themed exchange-traded fund and participate in a private technology investment vehicle. Although these investments appear diversified by company or asset structure, they may depend on similar assumptions about AI demand and infrastructure spending.  

Financial executives implementing AI internally face a related challenge. The companies supplying their technology may be financially strong today but dependent on continued fundraising to support future development. Vendor selection should therefore incorporate financial durability, contractual protections, data portability, service continuity and contingency planning.  

Is AI Still a Good Investment? 

The debate over AI investment increasingly resembles the argument surrounding the internet during the late 1990s. Optimists correctly anticipated that digital technology would transform the global economy. Skeptics correctly recognized that many companies attracted enormous valuations without developing sustainable business models. 

Both groups could be right again. 

AI's long-term economic significance does not guarantee attractive returns for every investor. An industry can transform society while destroying enormous amounts of capital through excessive competition, technological obsolescence, overbuilding and unrealistic valuations. 

Conversely, concerns about an investment bubble do not establish that AI companies are inherently overvalued. Businesses capable of generating substantial recurring revenue, maintaining competitive advantages and improving operating margins may justify valuations that initially appear extraordinary. 

The central question for investors is increasingly shifting from whether AI will succeed to which businesses will capture the economic benefits of that success. 

For financial professionals, this means evaluating AI investments with the same discipline applied to other assets. Revenue growth must be considered alongside profitability, cash generation, capital intensity, competitive positioning, dilution and valuation. Infrastructure investments require careful examination of utilization rates, financing costs, equipment life cycles and customer commitments. Private investments demand additional scrutiny of liquidity, governance and exit opportunities. 

Anthropic's historic public offering demonstrates that investor appetite for leading AI companies remains substantial. OpenAI's reported fundraising difficulties suggest that even exceptional businesses are encountering more demanding capital markets. Neither development conclusively establishes that the AI investment boom has peaked. 


Researched by DWN Staff

Written with assistance of ChatGPT

Loading suggested stories…