A company uses an AI to automatically summarize internal reports.

It is supposed to help managers recognize the most important information more quickly.

The AI sorts reports by relevance, urgency, and estimated impact.

Reports with positive developments usually appear at the top.

After a few months, however, something becomes noticeable:

Problems are often detected far too late.

The reports sound optimistic — but in reality, crises are increasing.

Question:
Why can an AI that efficiently sorts reports make bad news invisible?

Solution will be unlocked tomorrow.