The AI ​​party is over: $720 billion with no return and tech stocks are plummeting – is the billion-dollar bubble about to burst?

"Enough with chatbot games": SAP board member criticizes current AI strategy

$720 billion with no return: Is the tech world facing its biggest AI mistake?

For years, the hype surrounding artificial intelligence on the stock markets seemed to know only one direction: straight up. But in 2026, the tide turned dramatically. Tech giants like Amazon, Microsoft, and Google are investing hundreds of billions of dollars in building new data centers and powerful chipsets – and are suddenly reaping stock market crashes instead of cheers on Wall Street. Investors are growing uneasy and increasingly demanding tangible returns instead of vague promises for the future. Because while spending on AI infrastructure is reaching astronomical heights, the operating profits from these technologies are failing to materialize. Is the tech industry on the verge of a massive investment bubble bursting, or are we simply witnessing a much-needed return to economic reality? One thing is clear: the era of unlimited advance praise is over – the market now demands concrete results.

Now that the AI ​​party is over and the bill is coming due: Why Wall Street no longer trusts the AI ​​gold rush mentality

Throughout much of 2026, a pattern that has now taken on an almost ritualistic character repeated itself in the global capital markets. A major technology company releases its quarterly figures, simultaneously announcing a gigantic increase in its investment spending for the expansion of artificial intelligence, and within hours the stock markets react with a crash that sometimes wipes out several hundred billion dollars in market capitalization. What was considered a promise of the future for years is now increasingly seen as a risk factor. In the first week of February 2026 alone, the world's largest technology stocks lost more than a trillion dollars in value after Amazon, Alphabet, Microsoft, and Meta jointly announced investment plans of over 650 billion dollars for the construction of new data centers and AI infrastructure. This sum exceeds the economic output of entire countries such as Israel, Singapore, or the United Arab Emirates.

As the year progressed, skepticism intensified. In June 2026, a sudden sell-off in chip manufacturers and memory producers like Micron reignited widespread doubts about whether the massive expenditures on graphics processors, cooling systems, and energy infrastructure could ever be recouped through corresponding returns. The Philadelphia Semiconductor Index, a key indicator for the chip industry, plummeted more than 20 percent from its yearly high within a few weeks, crossing the technical threshold into a so-called bear market. In mid-July 2026, the pattern repeated itself when increased competition from China and growing doubts about the profitability of previous investments triggered a global sell-off in technology stocks, causing the Nasdaq index to plunge by more than 4 percent in a single week.

Multi-billion dollar bets with no visible return on investment

The core of the problem can be summarized as a simple imbalance. The four largest American technology companies—Alphabet, Amazon, Meta, and Microsoft—are planning capital expenditures of up to $720 billion for the current year, primarily for building new data centers, specialized AI chips, and the associated power infrastructure. This sum continues to grow quarter by quarter, while the actual revenue generated so far is disproportionately low. Amazon announced that its capital expenditures would rise to around $200 billion by 2026, more than $50 billion above analysts' expectations. The market reaction was unequivocal: Following this announcement, the stock initially plummeted by more than nine percent, and several research firms downgraded the stock from a buy to a neutral recommendation, citing not only the costs but also strategic risks to its core cloud and retail businesses.

Alphabet, Google's parent company, also came under pressure after announcing it could nearly double its investments this year to as much as $185 billion. Interestingly, the market reaction was initially more mixed than with Amazon, demonstrating that investors aren't inherently opposed to high spending, but rather to spending that isn't underpinned by a credible growth story for the cloud business. Once investors gain the impression that a company knows exactly what it's building capacity for and can utilize it effectively, even a drastic increase in spending is rewarded. Without this confidence, the same increase in spending is interpreted as a destruction of capital.

Tesla offers a particularly vivid example of this ambivalence. The electric car manufacturer announced it would double its investments this year to over twenty billion dollars to build capacity for artificial intelligence, humanoid robots, and autonomous vehicles. Since the company simultaneously reported quarterly profits and revenues above expectations, the stock remained stable despite the massive spending announcement, having even risen previously. This pattern illustrates that the stock market doesn't inherently penalize the volume of investment, but rather the gap between spending and demonstrable operational success.

Semiconductor industry as an early warning system for investors

A particularly sensitive indicator of sentiment surrounding artificial intelligence is the semiconductor industry, whose share prices are regularly among the first to react to doubts about the investment logic. In June 2026, memory chip manufacturer Micron experienced a share price collapse of more than thirteen percent in a single trading day, after its stock had risen by almost eight hundred percent the previous year, driven by the seemingly insatiable demand for memory chips for AI applications. Such an abrupt reversal after such a speculative rally is considered by many market observers to be a classic warning sign of a bubble approaching its breaking point.

Asian suppliers were not spared from the uncertainty either. South Korean memory chip manufacturers Samsung and SK Hynix each saw their share prices fall by around twelve percent within a short period, while the Taiwanese stock index TAIEX plummeted by more than six percent after contract manufacturer TSMC announced an additional investment of one hundred billion dollars in its American production facilities. Here, too, the same basic pattern emerges: An announcement intended as a growth signal is increasingly being interpreted by the markets as a warning sign of excessive capital commitment.

In July 2026, this trend continued when the Philadelphia Semiconductor Index plummeted by 10 percent in a single week, its worst weekly performance since April 2025. Even industry giants like Nvidia were not spared, temporarily losing so much value that their market capitalization briefly dipped below that of Apple, despite the two companies having vied for years for the title of the world's most valuable corporation. This development is remarkable because Nvidia had long been considered the clearest beneficiary of the AI ​​boom, essentially the company actually making money selling shovels in a gold rush while others merely invested in the mine.

The difference between infrastructure development and proof of yield

The crucial economic question underlying the current market nervousness concerns the time lag between capital commitment and return. Building a modern data center for artificial intelligence requires enormous upfront investments in land, buildings, cooling systems, network connections, and, of course, the graphics processing units (GPUs) themselves, whose prices remain high despite increasing competition. These investments must be made before a single additional dollar of revenue is generated from the corresponding capacity. As long as companies can credibly demonstrate that this capacity will be utilized and monetized in the foreseeable future, capital markets tolerate this lead time. However, as soon as doubts arise regarding utilization or pricing power for AI services, valuations plummet.

An additional risk is that an accelerated expansion of computing capacity could lead to a price collapse for AI services across the industry, because the supply of computing power is growing faster than the demand willing to pay. Analysts like Dan Ives of Wedbush have pointed out that the rapid expansion of global computing capacity is likely to put downward pressure on prices in the long run, thus reducing operators' returns on capital, which will ultimately necessitate a correction in investment dynamics. This concern corresponds to a classic economic pattern from previous technology cycles, such as the expansion of fiber optic infrastructure during the dot-com era, when massive upfront investments initially led to overcapacity and subsequently to a drastic price collapse before demand actually caught up years later.

Adding to the nervousness is a competitive factor. Increasing competition from China, particularly from cheaper and increasingly powerful AI models, challenges the premise that American technology companies possess a lasting technological lead that justifies their enormous investments. If Chinese providers can achieve comparable performance with significantly less capital, the entire investment strategy of Western hyperscalers is called into question, a fact immediately reflected in the share price reactions of July 2026, when intensified Chinese competition triggered one of the sharpest sell-offs of the year.

Our global industry and economic expertise in business development, sales and marketing - Image: Xpert.Digital

Industry focus areas: B2B, digitalization (from AI to XR), mechanical engineering, logistics, renewable energies and industry

More information here:

• Expert Business Hub

A thematic hub offering insights and expertise:

• Knowledge platform covering global and regional economies, innovation and industry-specific trends
• A collection of analyses, insights, and background information from our key areas of focus
• A place for expertise and information on current developments in business and technology
• A hub for companies seeking information on markets, digitalization, and industry innovations

Profit taking or genuine revaluation

Within the financial industry, there is no consensus on whether the recent price declines represent a fundamental reassessment of the AI ​​investment logic or simply overdue profit-taking after an exceptionally long rally. The technology sector within the S&P 500 had risen by almost 27 percent in the three months prior to the June 2026 sell-off, which many market strategists consider a natural trigger for consolidation, regardless of the fundamental valuation of the AI ​​story. Investment strategists like Brock Weimer of Edward Jones argued that, without a clearly identifiable fundamental trigger, the price declines should be interpreted as a technical correction following a strong rally since the March lows, rather than as an expression of a fundamental loss of confidence.

Other market observers take a more critical view. They point out that the valuations of leading AI companies have now reached a level that can only be justified by sustained high growth rates over many years, while at the same time the risk of disappointment increases with each additional quarter in which the promised productivity gains fail to materialize in the companies' balance sheets. This position is shared, among others, by investment experts who speak of a moment of risk reassessment, in which the market recognizes that while the AI ​​investment trend is not over, its valuation has been too far removed from the actual and foreseeable risks.

In this context, the differentiated investor reactions depending on the company are revealing. While Amazon and Microsoft were initially punished significantly for their investment announcements, Meta and, to some extent, Alphabet received a more positive response. This suggests that investors are not categorically opposed to high AI spending, but rather differentiate very carefully between companies that can demonstrate a convincing monetization strategy and those that are merely making upfront investments without showing a clear path to returns. This selectivity is economically sound because it directs capital allocation according to return prospects and not just based on fleeting hype.

Collateral damage in adjacent industries

Uncertainty surrounding the profitability of AI investments is not limited to the direct builders of data centers, but also affects suppliers and related industries. Software companies like ServiceNow and Salesforce have also come under significant pressure amid this general uncertainty, as investors increasingly fear that generative AI tools could displace traditional software solutions and their associated licensing models before the companies themselves have sufficiently benefited from the new technology. The software index within the S&P 500 lost almost eight percent in a single week in February 2026, with approximately one trillion dollars in market capitalization lost in this sub-segment during that period alone.

Data analytics companies like the Canadian firm Thomson Reuters and the British group RELX were also caught in the downward spiral, as investors fear that their business models, which rely heavily on proprietary datasets and editorial processing, could be increasingly challenged by powerful AI systems. RELX saw its share price plummet by more than 17 percent in a single week, the largest drop since the start of the COVID-19 pandemic in 2020. This development demonstrates that the debate surrounding AI investments is no longer solely a question of the capital costs for infrastructure operators, but increasingly affects the business models of established information service providers, whose value creation could potentially be disintermediated by generative systems.

In India, a key location for software exports and IT services, the uncertainty was also clearly evident. Shares of Indian software exporters lost an additional two percent within a single week in February 2026, following significant losses in the preceding weeks, bringing the total market value loss for this segment to more than twenty-two billion dollars. This illustrates the global reach of the debate, which is no longer just an American phenomenon confined to Silicon Valley, but has now affected the entire global technology supply chain.

The demand for an end to chatbot gimmicks

Within the corporate world itself, pressure is mounting to translate existing investments into concrete and measurable economic results. This was particularly evident in the pronouncements of Dominik Asam, CFO of the German software company SAP, who, in late July 2026, emphasized during the presentation of the quarterly results that artificial intelligence in enterprise software must finally move beyond the stage of simple chatbots and programming assistants. He described the vast majority of current AI language models as so-called low-hanging fruit—that is, simple use cases such as programming assistance or general chatbot dialogues, where occasional errors or so-called hallucinations of the system hardly matter because the risk of an incorrect outcome remains comparatively low.

The real challenge, according to Asam, lies in applying artificial intelligence to more complex and business-critical processes such as financial accounting, supply chain management, or other core business processes. Here, errors are amplified across multiple successive processing steps, meaning that even minor uncertainties at the beginning of a process can ultimately lead to significant and costly incorrect decisions. A single incorrect data point in an automated financial forecast or a flawed assessment in a multi-stage supply chain optimization can statistically accumulate over the entire process and ultimately generate substantial compliance and liability risks.

This assessment touches upon a key point in the current debate. Asam described the notion that a generally trained, generic large-scale language model can simply be laid over existing, often historically grown, fragmented enterprise data landscapes to solve all operational problems as fundamentally flawed. Instead, he argued, what is needed are specifically designed, governed systems that are tightly embedded in particular business processes and whose results can be reliably monitored. While this approach is more complex and involves significantly higher costs for the necessary computing power, the so-called token consumption, it offers, unlike general chat applications, genuinely robust business results.

This statement can be interpreted as a kind of reality check for the entire industry. While in recent years the sheer availability of increasingly powerful language models was considered a sufficient value driver, the focus is now clearly shifting towards the question of how these models can actually be reliably, controllably, and economically integrated into existing business processes. This shift from pure model performance to practical implementation quality is likely to become a key differentiator between successful and less successful AI strategies in companies in the coming years.

What this means for future investment logic

The current market correction should not be prematurely misinterpreted as the end of investments in artificial intelligence. Rather, it represents a long-overdue maturation phase in which the relationship between investors and borrowers is fundamentally changing. In recent years, the mere announcement of an AI project or a significant increase in the investment budget was often enough to trigger share price gains, because investors implicitly assumed that every additional investment would automatically lead to higher future returns. This assumption is now being systematically questioned, and companies will have to demonstrate much more concretely how their expenditures translate into measurable revenue and profit contributions.

For companies, this means a shift in their communication strategy with investors. Mere capacity announcements are no longer sufficient to generate positive share price reactions. Instead, key performance indicators (KPIs) such as data center utilization, the actual revenue contributions of individual AI products, the development of profit margins in the cloud business, and reliable forecasts regarding the amortization period of investments are moving to the forefront. Companies that can communicate these KPIs transparently and convincingly are likely to continue to be able to secure large investment budgets in the capital markets, while companies without a clear monetization strategy must expect continued valuation pressure.

This development presents both risks and opportunities for the overall economy. A too-abrupt decline in investment activity could have negative short-term effects on economic growth, as investments by large technology companies now constitute a significant share of overall investment activity in the United States. At the same time, disciplined capital allocation could lead to a healthier and more sustainable development of the industry in the long run, with capital flowing more strategically into projects with demonstrable economic benefits, rather than into an undirected and sometimes speculative expansion of computing capacity. The upcoming quarterly reports from the major technology companies will be crucial in determining whether this disciplining effect of the capital markets prevails or whether the structural competitive dynamics among the hyperscalers continue to drive ever-new investment records, regardless of the short-term market reaction.

☑️ Our business language is English or German

☑️ NEW: Correspondence in your native language!

Konrad Wolfenstein

I and my team are happy to be available to you as your personal advisor.

You can contact me by filling out the contact form here [email protected]:or simply call me at +49 7348 4088 965. My email address is

I'm looking forward to our joint project.

☑️ SME support in strategy, consulting, planning and implementation

☑️ Creation or realignment of the digital strategy and digitization

☑️ Expansion and optimization of international sales processes

☑️ Global & Digital B2B trading platforms

☑️ Pioneer Business Development / Marketing / PR / Trade Fairs

From visibility to trust: Your scalable path with Xpert.Digital - Image: Xpert.Digital

In industrial B2B, sustainable business relationships rarely emerge overnight. They develop step by step – through visibility, professional relevance, recurring touchpoints, and growing trust. Xpert.Digital's 4-stage model addresses precisely this: It offers a structured path that begins with a manageable entry point and can evolve into deeper collaboration in business development if needed.

Instead of relying on loud marketing promises, this model puts the relationship at the forefront. Companies start with clearly defined, easily calculable measures and then decide, based on their own experience, how far they want to expand the collaboration. A key factor for this undisturbed trust-building process: The platform completely avoids annoying advertising ads, so the editorial focus remains solely on the companies' expertise.

More information here:

• From visibility to trust: Your scalable path with Xpert.Digital