For developer-facing, near-term hardening of AI agents, AAISM prioritizes secure agent design and runtime controls against prompt injection, unsafe memory/tool use, and tool-execution compromise. These are primary exploitation paths for agents that read external content, persist memory, and call tools with elevated privileges. Training must center on: guarding tool invocation, constraining memory scope, sanitizing
/validating inputs, and isolating high-risk actions. Topics like bias/fairness (B) and policy/hallucinations (C) are important but are governance/assurance concerns; API abuse and plug-in risk (D) matter, yet the core, developer-controlled attack surface for agents is injection and unsafe tool/memory design.
References:* AI Security Management (AAISM) Body of Knowledge: Agent Security-prompt injection defenses, tool execution constraints, memory governance* AAISM Study Guide: Developer controls for agent frameworks; input validation, sandboxing, scoped permissions, guardrail patterns