Risks from lack of robustness in AI systems including vulnerability to adversarial attacks, distribution shifts, and unexpected inputs.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.