In PMI-CPMAI, persistent issues like data misuse across departments and jurisdictions point directly to weaknesses in AI and data governance, not just ethics awareness. While ethics guidelines are important, they are only one element of a complete governance framework. PMI’s AI governance view stresses the need for a detailed, actionable governance strategy that defines roles (owners, stewards, custodians), access controls, data classification, data use policies, approval workflows, and compliance processes that consider regional regulations (e.g., differing data protection laws).
Without such a governance plan, teams may unintentionally share or use data in ways that conflict with internal policies or external regulations, even if they know and care about ethics. Algorithmic bias (option C) and explainability (option A) are important but do not directly address cross-department access management and regional regulatory differences. Failure to implement robust encryption (option D) concerns technical security of data in transit/at rest; it does not, by itself, prevent misuse by authorized but improperly governed users.
Therefore, the root issue causing these challenges is the lack of a detailed plan addressing a governance strategy (option B), which should integrate ethics, regulatory requirements, and operational controls for data use across departments and regions.
====================
