An AI project team is in the process of designing a security plan.

In PMI-CPMAI, security planning for AI solutions goes beyond traditional technical controls; it explicitly includestransparency, explainability, and regulatory complianceas part of a responsible AI posture. The guidance states that security and trust in AI depend not only on encryption, access control, and infrastructure hardening, but also on whether stakeholders can understand how decisions are made and whether those decisions comply with applicable laws and policies.

PMI’s AI management perspective includes requirements forexplainable and auditable decision-making, particularly in public-sector and high-impact domains. This means designing systems so that model behavior can be interpreted, key features and factors identified, and decisions documented in a way that regulators, auditors, and affected users can review. The project manager is therefore expected to ensure that the AI system’s design and governance support transparency and explainability, in addition to technical security controls.

Focusing only on technical measures or assuming compliance without review contradicts PMI-CPMAI’s emphasis on proactive governance and legal/ethical due diligence. Reliance solely on encryption addresses confidentiality but not fairness, accountability, or understandability. Thus, the correct action is toensure the AI system’s decisions are transparent and explainable, embedded alongside other security and compliance safeguards.