Why good AI fails: the missing layer in digital transformation
- Emiel Kruijd
- Jun 27
- 2 min read
Updated: Jul 30
AI pilots often fail — not because the code is wrong, but because the people aren't ready.
Why do AI pilots fail even when the technology works? Because successful digital transformation is not just about algorithms — it’s about adoption, human behaviour, and organisational readiness. In this blog, we explain why a promising healthcare pilot failed, and how AIM helps organisations avoid the same pitfalls using our AI Readiness Scan and our Seven Transformation Domains.
A crisis, the pressure, the promise
In the middle of the COVID-19 pandemic, hospitals were stretched to the limit. Reducing physical contact was essential. Efficiency wasn’t a luxury — it was a necessity. So, when we launched a pilot with smart cameras to measure heart rate and body temperature remotely, it seemed like the perfect solution:
No wires
No contact
Reliable, real-time data
Less workload for staff
Technically, everything worked. In practice, it didn’t.
Patients didn’t trust it. Even after clear explanations that no video footage was stored — only anonymised data — patients felt watched. The mere presence of a camera in their room was enough to create discomfort. In a high-stress, vulnerable context like healthcare, perceived surveillance is a serious barrier.
Nurses didn’t embrace it. What the system replaced wasn’t just a task — it was a moment of human connection. A hand on the forehead. A short conversation. Subtle cues that shaped clinical intuition. Removing that touchpoint removed trust in the data.
The result?
Most nurses continued taking manual temperature readings, just to be sure. Instead of saving time, the system added complexity.
The real insight: this wasn't a tech failure — it was a transformation failure
AI didn't fail because the algorithm was wrong. It failed because the context wasn't ready.
At AIM, we see this mistake often. Organisations push pilots forward because “we need to do something with AI” — without asking the harder question: Do our people understand it, trust it, and see its value in their work?
That’s why we approach AI differently. Before we build or pilot anything, we start with a Digital Blueprint Scan to assess organisational readiness. This ensures people understand the ‘why’, and helps leaders embed AI in daily practice.
Want to avoid these pitfalls? Explore our Seven Transformation Domains.
