Threat vectors that companies with ML need to consider

The image above shows the threat vectors that companies that are training their own ML models need to consider.

Note that this diagram assumes several things about the ML ecosystem including: the company is training its own ML model, the ML model is customer-facing.