Organizations increasingly seek to integrate generative AI into their workflows, yet deciding which tasks to augment, and to what degree, remains a poorly structured challenge. Current design toolkits such as Microsoft’s HAX and Google’s PAIR offer valuable guidance for building human-AI interfaces but assume the decision to deploy AI has already been made. This paper addresses the upstream question: is a given task ready for AI augmentation? We present a five-dimensional framework that evaluates tasks along Cognitive Load, Temporal Engagement, Expertise Requirement, Social-Relational Demands, and Agency & Responsibility. The first three dimensions capture readiness—malleable conditions where AI can reduce burden and extend human capability—while the latter two capture suitability—structural and normative constraints that may limit AI involvement regardless of technical feasibility. Each dimension is operationalized through four concrete factors, grounded in a systematic mapping of 41 HAX and PAIR guidelines. We validate the framework’s content through an expert survey (n = 10), which confirmed strong relevance and clarity across all dimensions and factors, and a focus group with practicing AI consultants (n = 5). The framework aims to serve as a structured precursor to design-level toolkits.