New external attack methods against AI systems and services appear daily; confirm the points to watch in order to respond to these risks.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.