Project-based organizations are increasingly exposed to a combination of uncertainty, accelerated technological change, and rising sustainability expectations. In this context, many organizations are investing in digital platforms, analytics capabilities, and artificial intelligence applications, yet they often struggle to transform dispersed project information into reliable and actionable knowledge. The spread of generative AI has intensified this tension: while organizations can now generate answers, summaries, and recommendations at scale, they frequently lack adequate governance mechanisms for knowledge provenance, accountability, privacy, and ethical use. Starting from the INSPIRE PM framework and its broader grounding in Project Management Evolutive (PME), this paper argues that sustainable project performance depends on the ability to build Human–AI Hybrid Knowledge Ecosystems (HA-HKE), that is, socio-technical configurations in which human actors and AI systems collaboratively generate, validate, circulate, and apply project knowledge under explicit governance rules. The paper develops a conceptual architecture for such ecosystems, identifies the key knowledge assets and flows required across the project lifecycle, and proposes a design-oriented research agenda centered on three artifacts: a hybrid knowledge ecosystem blueprint, a decision intelligence loop, and a lightweight governance checklist for responsible GenAI use in project contexts. The paper contributes conceptually by reframing project management as a knowledge ecosystem challenge, methodologically by outlining a design science approach with multiple-case evaluation, and managerially by offering a practical pathway for PMOs and project leaders seeking to combine AI-enabled decision support with sustainability, traceability, and human accountability.