I wanted to start a broader practical use cases thread around AI Scam & Fraud Detection.
A useful angle is limitations, but it also connects with customisation. Different tools and workflows can produce very different results even when the goal is similar.
It would be interesting to compare...