How to Use AI the Right Way
- And Mistakes to Avoid
Artificial Intelligence is no longer a futuristic promise — it’s an everyday business tool. Yet many companies still struggle to implement it successfully.
At the AI Summit Kitzbühel 2025, several speakers offered a clear message: AI won’t transform your business unless you use it the right way. And often, the biggest obstacles aren’t technical — they’re human.
So, how do you ensure your AI initiatives don’t fail before they even start? Let’s explore what not to do, and how to build AI systems that truly support your team, your goals, and your customers.
Mistake #1
Starting with Tools, Not People
One of the clearest lessons from the AI Summit Kitzbühel was this: AI transformation isn’t about tools — it’s about people. IBM’s opening keynote put it perfectly: real transformation starts with mindset, not machines. Deborah Hüller and Daniel Unkelhäußer shared how IBM became its own “Client Zero,” resolving 94% of HR requests with AI while still focusing on human-centered design.
Do this instead: Start with your employees’ real pain points. What repetitive tasks drain time? Where is decision-making slow? AI should help, not overwhelm. Begin with pilot projects and co-create with your team. Don’t “roll out” AI — test it together.
Mistake #2
Assuming Everyone Must Be a “Prompt Engineer”
The push for everyone to master prompt engineering can backfire. Lena Kornau highlighted this with a pointed question: “If you’ve drunk a glass of wine, does that make you a winemaker?” Expecting non-technical employees to speak fluent AI jargon leads to frustration, fear, and disengagement.
✅ Do this instead: Train people based on their needs and roles. Not everyone needs to be a power user. Simple, clear instructions, practical use cases, and relatable language go much further than buzzword-heavy workshops.
Mistake #3
Hiding Behind AI Buzzwords
Throwing around terms like “LLMs,” “GPUs,” or “data silos” may impress your board — but they won’t help your team understand what’s changing. In fact, this kind of language often leads to resistance. People don’t reject AI because they fear technology — they fear confusion.
✅ Do this instead: Communicate honestly. Explain what AI can do, and what it can’t. Be transparent about changes, timelines, and expectations. Avoid the fairy tale of “nothing will change” — because it will.
Mistake #4
Replacing, Instead of Supporting, Human Work
Olga Calvo emphasized that generative AI must be embedded into services that augment human capabilities, not just automate for the sake of speed. At Amazon, GenAI is used to enhance how sellers list products, how they process information, and how analysts summarize sports matches — always with humans in the loop.
✅ Do this instead: Use AI as a personal assistant, not a replacement. Where can AI remove “undifferentiated heavy lifting” so people can focus on the work they’re best at?
Mistake #5
Treating AI as a One-Time Project
Real AI success takes iteration. Amazon, IBM, and others shared one common strategy: start small, learn fast, scale wisely.
AI needs clear goals, feedback loops, and the freedom to evolve with your business. If you treat it as a fixed solution, it will likely fall short.
✅ Do this instead: Build prototypes, test with real users, and refine as you go. Lena Kornau’s advice? Make room for creativity — don’t assume AI can’t support innovation. In fact, with the right prompts and feedback, AI can be your most surprising idea generator.
The Bottom Line: Get the Culture Right
Across every session — from IBM’s keynote to Lena Kornau’s masterclass — the message was consistent: AI transformation is 80% cultural, 20% technical.
Trust, curiosity, and communication are the real drivers of success.
If your team understands AI, feels part of the process, and sees real benefits, adoption will follow. But if you skip the human element? Even the best AI won’t deliver.
Looking Ahead
The next AI Summit Kitzbühel will take place in June 2026. Until then, the momentum continues — with companies already implementing ideas discussed at the summit, and new collaborations forming across sectors.
For those who couldn’t attend — or want to relive it:
Sources:
- IBM Keynote: “The Future of Work: powered by AI, driven by people”
- Lena Kornau: Don’t Do This!
Wie man KI garantiert falsch einführt” - Olga Calvo: Building with AI at Amazon:
From Customer Obsession to Generative Innovation” - TechWellTold: “After Deployment comes the frustration
Why AI implementations fail…”