Research into two-party agentic delivery
Agent identity, permissions, prompt injection, privacy, data meaning, memory, technical dependencies, licensing, and fit-for-purpose service architecture.
AAO-014 / Data, Technology, and Security
Infrastructure managers should authorize provider agents through task-specific identities and tested revocation. Initial use should remain limited to approved low-risk support, with operational engineering defining how access can be withdrawn safely.
AAO-012 / Data, Technology, and Security
Vendor privacy managers should trace personal information through the stores and tools an agent uses. A processing schedule becomes operational only when both firms can locate information, apply instructions, and evidence correction or deletion where required.
AAO-002 / Data, Technology, and Security
Outsourcing security managers should test whether untrusted supplier content can cause a trusted agent to act outside its mandate. The test must follow content through retrieval, memory, tool selection, and the final business action.
AAO-003 / Data, Technology, and Security
Executives should approve a processing activity map before delegating work to provider agents. Responsibility follows the actual purposes and decisions in the chain, while contractual labels alone cannot establish who controls personal information.