Why 80% of AI projects do not deliver ROI

Why doesn't 80% of AI projects deliver ROI?
This question is increasingly heard not from technologists, but from management and company owners.

In practice, the problem rarely lies in the quality of models or data.
The lack of systemic responsibility is much more common:
who makes decisions, who supervises them and who is responsible for the consequences of AI's action in real trials.

Without clear governance, integration with processes and organisational readiness
AI remains an experiment — even if technologically "works".

This post accurately shows that ROI in AI is a test of the maturity of the organization rather than the technology itself.

Source: LinkedIn — IZOB