Why Do 95% of Gen-AI Projects Fail? The Deadly Trap of Hype

Session Overview

Around the world, countless companies are launching generative-AI projects—but only about 5% translate into real revenue growth. In other words, 95% miss expectations. The root cause isn’t the technology itself; it’s the absence of an “execution DNA.”
Drawing on hands-on experience building AI agents for banks and retailers, this session identifies common failure patterns and presents practical strategies to overcome them.

We’ll cover four pillars:

• Business goals & ROI reality check — Start from the business, not the tech.
• LLM strengths and limits — Balance imagination with rigorous risk management.
• Tool use & data readiness — Poorly prepared data is the fastest path to failure.
• The wall of cultural transformation — Organizational change and execution capacity.