What remains? Three years after the ChatGPT hype: The grand AI dream meets economic reality

The disillusionment after the euphoria: When hype promises meet management reality

Three years after the "ChatGPT moment," disillusionment has set in at the executive level. While tech giants like Meta and Google continue to invest hundreds of billions of dollars in AI infrastructure, the broader business world presents a different picture: stagnation instead of revolution.

ChatGPT was released to the public by OpenAI on November 30, 2022. The system achieved record speeds in user acquisition and is considered the trigger for the massive AI hype that swept through the business world from 2023 onwards.

It was supposed to be the biggest productivity boost in history. But three years after the global hype surrounding generative AI, a dangerous gap has opened up between technological promise and economic results. Recent data from Forrester and the Boston Consulting Group paints a picture of "expensive stagnation": Only a vanishingly small percentage of companies have so far been able to translate their immense investments into real added value.

The case of fintech giant Klarna, in particular, serves as a warning shot for the entire industry. What was celebrated as a triumph of efficiency—the replacement of 700 employees with AI—turned out to be a boomerang for customer satisfaction. The lesson is painful, but necessary: ​​technology without empathy and strategic change management may save costs in the short term, but it destroys customer relationships in the long run.

This article looks behind the glossy press releases. We analyze why 2026 will be the year of major AI corrections, why the "cultural component" is the real killer of AI projects, and why technology alone cannot replace a missing corporate strategy. An assessment of the landscape between billion-dollar bets and the return to economic common sense.

The core problem: Reality meets expectation

The discrepancy between invested capital and realized returns is alarmingly clear. A Forrester study from 2025 shows that only 15 percent of surveyed executives were able to significantly improve their profit margins through AI implementations. This is not a fringe phenomenon or a problem limited to startups. It affects the entire economy, from the most financially powerful corporations to medium-sized organizations. Even more dramatic is the finding of the Boston Consulting Group: a mere 5 percent of surveyed executives reported widespread value creation effects from AI. This is not the definition of transformative change. It is the definition of stagnation despite expensively acquired infrastructure.

These figures become even more significant when viewed in the context of such expenditures. Meta alone announced investments of $70 to $72 billion for 2025, with a forecast of $600 billion by 2028. Google plans to invest $91 to $93 billion in 2025. Microsoft is also continuously increasing its AI capital budget. These are not investments in side projects, but core investments intended to define the future competitiveness of these companies. However, while the tech giants are forging ahead with unprecedented sums, a contrasting trend is emerging among companies outside this technological "inner circle": strategic delay.

Forrester predicts that roughly a quarter of planned AI investments will be postponed in 2026. This isn't about cutting speculative spending for cost reasons, but rather about postponing strategic projects that were high on the agendas of CFOs and CEOs because return on investment (ROI) expectations weren't met. A quarter of planned investments—that's not just a decline, but a systemic reassessment of the strategic importance of this technology.

The Klarna case: A warning in the form of a case study

The case of the Swedish fintech company Klarna is instructive here – not because it is an isolated incident, but because it vividly illustrates the systemic problem. In 2023, Klarna made international headlines with the announcement that it would replace 700 customer service employees with an AI chatbot system developed in collaboration with OpenAI. The figures were impressive: The chatbot handled two-thirds of all customer inquiries, was fluent in over 35 languages, and reduced response times from an average of 11 minutes to approximately 2 minutes. This is undoubtedly a remarkable operational achievement.

But by 2024, the underlying problems had already become apparent. Customer satisfaction had plummeted by 22 percent. This wasn't a statistical inaccuracy, but a clear signal from users that the system was reaching its structural limits. The AI ​​chatbot could handle simple transactional inquiries, but it was systematically overwhelmed by more nuanced issues—situations that required an understanding of the specific context, emotional intelligence, and, above all, empathy. When CEO Sebastian Siemiatkowski admitted the mistakes in 2025, his analysis was remarkably clear: The one-sided focus on cost efficiency had led to a decline in quality. In other words, the technology had been optimized to improve internal metrics, but not designed to ensure the actual customer experience.

The response was logical: In 2025, Klarna began rehiring customer service representatives and established a hybrid model where AI handles routine inquiries and human agents resolve complex cases. While a calculated savings of $60 million was maintained, overall customer service costs tended to rise again, as both the AI ​​infrastructure and a substantial human staff now had to be maintained. This is not a success story of automation, but rather an expensive lesson about the limitations of technical optimization without strategic change management.

The organizational dimension of failure

The core problem lies not primarily in the technology itself, but in the organizational ability to integrate it effectively. Research on change management shows that approximately 70 percent of all transformation initiatives fail to meet their objectives. This rate is even more pronounced in AI-specific projects: estimates suggest failure rates of 80 to 95 percent if companies do not establish clear goals, defined metrics, or consistent management frameworks.

The reasons for this failure are structural, not technical. First, there is a significant trust gap between management and staff. Studies show that 50 to 70 percent of employees express fear of profound technological changes. This fear is not irrational, but based on legitimate questions: How will my job change? Will I lose status or expertise? Will the work be done in addition to my existing responsibilities without providing me with resources or recognition? Leaders tend to underestimate these questions or interpret them as resistance to progress, rather than understanding them as systemic implementation problems.

Secondly, there is a fundamental gap between management's strategic intentions and operational feasibility. Fewer than 30 percent of companies with AI initiatives have established defined adoption metrics. This means that most companies are introducing AI systems without clearly defining what successful adoption actually means or how to measure progress. It's comparable to a construction project without blueprints or quality controls. The technology is implemented because it is considered strategically necessary (“fear of missing out”), not because there is a clear expectation of benefit.

Third, significant data problems are emerging that cannot be solved simply through investment. 73 percent of organizations cite data quality or data accessibility as their biggest challenge. This is not a question of technological resources, but of organizational maturity. Companies that have organized data in silos for decades cannot simply break down these structures by introducing an AI system. The result: AI systems work with low-quality inputs and consequently produce low-quality output ("garbage in, garbage out").

The limits of automation: The customer experience paradox

Another phenomenon is clearly evident in the automation of customer service. ServiceNow reports that AI systems are capable of autonomously handling approximately 80 percent of simple customer inquiries. Resolution times can be reduced by 52 percent and first contact resolution rates improved by 40 percent. These are impressive operational metrics. However, customer studies simultaneously show that 93 percent of customers prefer a human contact person for complex issues. This is not a matter of personal preference, but reflects a fundamental limitation.

Most real-world customer problems are not simple. They are context-dependent, often emotionally charged, and require an understanding of the individual situation. A customer experiencing difficulties with a refund needs not only a quick response but also the feeling of being understood. With complex financial products, the customer needs to trust that their counterpart is looking out for their interests. These are qualities that are fundamentally beyond the reach of mechanical automation because they require judgment and genuine human connection.

The data suggests that AI systems in customer service are most effective when they act as tools for human agents (a "co-pilot"), not as replacements. A system that supports employees with routine tasks, automates documentation, or pre-researches information yields positive results. A system that attempts to completely replace humans often leads to a chain of dysfunctional effects: customers switch providers, complaint rates increase, and brand trust declines. The operational goal of cost reduction is thereby undermined because customer churn and reputational damage are more expensive than the savings achieved.

A new dimension of digital transformation with 'Managed AI' (Artificial Intelligence) – Platform & B2B Solution | Xpert Consulting - Image: Xpert.Digital

Here you will learn how your company can implement customized AI solutions quickly, securely, and without high entry barriers.

A Managed AI Platform is your all-round, worry-free package for artificial intelligence. Instead of dealing with complex technology, expensive infrastructure, and lengthy development processes, you receive a turnkey solution tailored to your needs from a specialized partner – often within a few days.

The key benefits at a glance:

⚡ Fast implementation: From idea to operational application in days, not months. We deliver practical solutions that create immediate value.

🔒 Maximum data security: Your sensitive data remains with you. We guarantee secure and compliant processing without sharing data with third parties.

💸 No financial risk: You only pay for results. High upfront investments in hardware, software, or personnel are completely eliminated.

🎯 Focus on your core business: Concentrate on what you do best. We handle the entire technical implementation, operation, and maintenance of your AI solution.

📈 Future-proof & Scalable: Your AI grows with you. We ensure ongoing optimization and scalability, and flexibly adapt the models to new requirements.

More about it here:

• The Managed AI Solution - Industrial AI Services: The key to competitiveness in the services, industrial and mechanical engineering sectors

Reality check: Who really benefits from AI today?

The available data suggests a split in the economy. On the one hand, there are the tech giants and a few specialized "AI-native" companies that continue to invest heavily in AI infrastructure and deeply integrate it into their business models. On the other hand, there is the vast majority of traditional companies that have adopted AI but see only limited success in terms of value creation.

McKinsey data shows that around 23 percent of companies are actively scaling AI systems, while 39 percent are still in experimental phases. This means that while 62 percent are engaged with AI in some way, their commitment is by no means homogeneous. Companies with clear AI strategies and established governance structures achieve a roughly 2.5 times higher ROI than those that implement AI ad hoc or as a purely tactical initiative. The top performers, achieving a tenfold return on investment, are an exclusive group. These are companies that understand AI not as an isolated IT solution, but as an integrated component of a comprehensive business transformation.

BCG reports that the average ROI is currently 11.2 percent, while mature organizations are already achieving returns twice as high. This is not a trivial difference. It means that organizational maturity is two to three times more important than pure technological capability. By comparison, a traditional business focused on operational efficiency can expect a 15 to 20 percent return. AI initiatives, therefore, are not competing on a level playing field; they must deliver exceptional returns to justify the technology's inherent risks.

The investment paradox: More money, less trust

The phenomenon emerging for 2026 is remarkable. While tech companies continue to invest record sums in AI, trust among traditional businesses is declining. Meta, Google, and Microsoft are drastically increasing their budgets. Yet, at the same time, traditional companies are recalibrating their AI plans.

Forrester predicts that 25 percent of planned AI investments will be postponed until 2027. This is not a retreat, but a replanning. The message from companies is clear: "We will invest in AI, but only when we clearly see the benefits." This marks the transition from a phase of speculative experimentation to a phase of results-oriented investments.

A second phenomenon exacerbates this dynamic: measurement blindness. 46 percent of companies have not established a structured framework for measuring ROI. This means that almost half of the investing companies don't really know if their projects are working. Considering that an average AI initiative takes three to five years to reach full value, this leads to a scenario in which companies allocate budgets for years without having valid metrics for success. It's like driving in complete darkness – hoping to eventually reach the destination.

The cultural component: The deep organizational problem

Herein lies the real problem. AI implementations don't fail because the technology fails. They fail because companies try to apply technological solutions to organizational problems that are cultural in origin. Studies indicate that cultural factors and resistance are the primary barriers in over 50 percent of failed AI initiatives.

This manifests itself on several levels. First, there is widespread fear of job loss. Companies implementing AI rarely communicate openly that the technology could replace roles. They talk about "automation," "efficiency," and "productivity." But employees understand the subtext. If this fear is not addressed through genuine retraining, clear role definitions, and job guarantees, it leads to covert resistance, low acceptance, and a kind of passive refusal.

Secondly, there is a fundamental trust issue with AI systems themselves. Many employees are skeptical of AI's ability to make nuanced decisions. They worry about bias, false positives, and the risk of automated systems overlooking important context. This skepticism is not unfounded. There is ample evidence of hallucinations in AI models and error-proneness in special cases that are underrepresented in the training data. If employees don't understand how an AI arrives at a decision, they will either ignore the system or lose trust in the organization itself.

Third, structural deficiencies are revealed. Organizations with deep functional silos cannot effectively utilize AI systems designed for cross-functional collaboration. Companies whose evaluation systems prioritize individual performance over collaboration will struggle to invest in collaborative AI models. Middle management, feeling threatened by automation, will erect subtle barriers to adoption. These problems cannot be solved with better software, but only with genuine organizational redesign.

The lesson: Technology is no substitute for strategy

From all this data, one lesson emerges that is not new, but needs to be relearned in this context: Technology alone does not solve business problems. It is a tool. A powerful tool in the hands of organizations that know how to use it – and a very expensive toy in the hands of those who hope for magical change.

Companies that make real progress with AI do several things in parallel: They have a clear business strategy in which AI plays a specific role, rather than being the all-encompassing solution. They invest in change management with the same energy and budgets as they do in the technology itself. They establish clear measurement frameworks before implementation. They continuously train their employees to work in an AI-enhanced environment. They proactively address cultural resistance. And they establish strong governance structures to ensure that AI systems align with the company's values.

These are not simple or quick processes. Deloitte research shows that “agentic AI”—the next wave of AI—takes an average of three to five years to deliver real added value. This is not a criticism of the technology, but a realistic understanding that deep organizational transformation takes time.

The drifting apart: Who wins and who loses?

A fascinating phenomenon emerges when considering who has successfully implemented AI. Meta, Google, and Spotify continue to invest heavily and report positive results. These are companies with a deep understanding of data science, an established culture of innovation, and the resources to tolerate mistakes and learn from them. Klarna, on the other hand, introduced AI primarily for cost reasons, overlooking the strategic dimension.

This outlines the contours of a two-tier economy. The first group consists of companies that understand AI as a transformative tool and possess the necessary structures, data, and cultures. The second group comprises traditional companies that want AI because their competitors are doing it, but lack the organizational maturity. This group will continue to experiment, spend money, and achieve limited success, while accumulating structural competitive disadvantages compared to the first group.

This dynamic will intensify over the next five years. Organizations that invest in change management and organizational maturity alongside their technology investments now will be the winners. Those who invest solely in technology and hope for automatic transformation will fail.

Outlook: 2026 and beyond

Forrester's prediction for 2026 is spot on: "The art of the possible gives way to the science of the practical." The era of speculative experiments is coming to an end, and the era of results-oriented investments is beginning. CFOs will be involved in AI decisions not out of enthusiasm, but because they have clear return expectations. The fact that 30 percent of large companies will introduce mandatory AI training indicates an acceptance that organizational competence still needs to be developed. Companies that postpone their AI plans are no longer seen as losers, but as prudent, because they realistically assess the time and organizational requirements.

The message for business leaders is clear: The AI ​​hype isn't over. The technology is real and will continue to deliver results where traditional systems fail. But the naive belief that AI investments alone will deliver transformative outcomes is a thing of the past. The next phase of AI adoption will be defined not by technological, but by organizational breakthroughs. Those who understand this will win. The others will waste years and capital, only to end up where they should have started: with a strategic, integrated, and human-centered approach.

☑️ Our business language is English or German

☑️ NEW: Correspondence in your national language!

Konrad Wolfenstein

I would be happy to serve you and my team as a personal advisor.

You can contact me by filling out the contact form or simply call me on +49 89 89 674 804 (Munich) . My email address is: wolfenstein ∂ xpert.digital

I'm looking forward to our joint project.

☑️ SME support in strategy, consulting, planning and implementation

☑️ Creation or realignment of the digital strategy and digitalization

☑️ Expansion and optimization of international sales processes

☑️ Global & Digital B2B trading platforms

☑️ Pioneer Business Development / Marketing / PR / Trade Fairs

Benefit from Xpert.Digital's extensive, fivefold expertise in a comprehensive service package | R&D, XR, PR & Digital Visibility Optimization - Image: Xpert.Digital

Xpert.Digital has in-depth knowledge of various industries. This allows us to develop tailor-made strategies that are tailored precisely to the requirements and challenges of your specific market segment. By continually analyzing market trends and following industry developments, we can act with foresight and offer innovative solutions. Through the combination of experience and knowledge, we generate added value and give our customers a decisive competitive advantage.

More about it here:

• Use the 5x expertise of Xpert.Digital in one package - starting at just €500/month