The challenge is proving greater than imagined. AI is developing rapidly, expectations are high, and the pressure to participate is great. At the same time, there is often a lack of knowledge and experience to deploy AI in a thoughtful way. As a result, many organizations apply it within the same structures, processes and ways of thinking as always. And as the saying goes, "If you do what you did, you get what you got."
Companies are taking a step back
From the moment AI became widely accessible, companies could not embrace it fast enough. Some with even greater ambitions than others: speed up processes, reduce costs, or even replace entire departments with AI. For a while, nothing seemed too crazy. But now we are seeing a turnaround. A well-known example is Klarna. In 2023, the company had many marketing tasks performed by AI, and by 2024 AI was conducting two-thirds of customer service calls. Not long after, Klarna moved away from that, due to noticeable loss of quality (1).
And Klarna is not the only example. According to Gartner (2,3), more and more pioneers are withdrawing their grand AI plans. The reasons are recognizable: costs are rising, security turns out to be more complicated than thought, and integrating AI into existing processes demands more from organizations than they had anticipated. Moreover, in practice AI turns out to be less empathetic than a human.
Add to this the figures from MIT (4) - only 5% of AI projects actually deliver value, while the rest often remain stuck in the pilot phase - and it becomes clear that many organizations are currently struggling to realize their ambitions.
This raises one important question: How do we ensure that AI does deliver that promised value?
The mindset for success
According to digital analyst, anthropologist and award-winning author Brian Solis, success with AI begins not with the technology itself, but with the way an organization thinks, learns and adapts to new possibilities (5).
He advocates a fundamental mindshift: "Stop thinking in terms of automation and cost savings, and start creating real new value."
Solis identifies three key shifts:
- From optimizing to innovating: Not "how can we do this faster?" but "what can we do completely differently now?"
- From control to creation: Give room for experimentation and learning, rather than tightly defining everything in advance.
- From inside to outside: Focus on the customer and the value you deliver, not on your existing internal processes.
This approach requires new leadership - curious, learning and open to collaboration between man and machine. It is precisely this mindset that distinguishes companies that do succeed with AI, and helps them see beyond the hype.
How can you do it differently?
AI only delivers value if you use it intelligently and consciously. This does not start with more technology, but with daring to change.
- Choose sharply: Start with one concrete use case that really adds value - starting small is powerful if the goal is clear.
- Redesign: Use AI not to speed up existing processes, but to reinvent the work itself.
- Invest in people: Technology only works when people understand, trust and empower it. Combine technical knowledge with change power.
- Small steps, big impact: Iteratively test, learn, improve and only then scale up.
- Measure what matters: Focus not just on efficiency, but on real impact - customer satisfaction, quality, speed, growth.
In short: Approach AI not as an IT project, but as a transformation of how you work, learn and create value.
From replacing to innovating
Klarna's examples and Gartner's analyses show that organizations that simply replace people (or existing processes) with AI often hit their limits. Not because AI is failing, but because the work itself has not been reinvented. Those who use AI as a driver of innovation - and not as a cheap substitute - discover the real value: faster decision-making, better collaboration and greater innovativeness. Because one thing is certain: If you do what you did, you get what you got. Want to achieve something different? Then you have to dare to do something different.
- Futurism: https://futurism.com/klarna-openai-humans-ai-back
- Gartner: https://www.gartner.com/en/newsroom/press-releases/2025-06-10-gartner-predicts-50-percent-of-organizations-will-abandon-plans-to-reduce-customer-service-workforce-due-to-ai
- Gartner: https://www.gartner.com/en/newsroom/press-releases/2025-06-25-gartner-predicts-over-40-percent-of-agentic-ai-projects-will-be-canceled-by-end-of-2027
- Fortune: https://fortune.com/2025/08/18/mit-report-95-percent-generative-ai-pilots-at-companies-failing-cfo/
- Brian Solis: https://briansolis.com
