From Bureaucracy to Algorithm: How Multilateral Development Banks Navigate AI Implementation without Losing Institutional Legitimacy

Multilateral Development Banks face a critical paradox: they must adopt artificial intelligence to address the $2.5 trillion annual development financing gap while maintaining legitimacy across stakeholder ecosystems demanding human oversight and cultural sensitivity. Through analysis of 30 MDB staff members and 31 diverse stakeholders, we reveal that 44% of MDBs lack formal AI frameworks—not due to organizational failure, but as rational Legitimacy-Centered strategy. While 56% of staff identify technical complexity as their primary barrier, stakeholders prioritize governance (90% demand human oversight) and cultural adaptation (90% require local language support) over technical sophistication. We develop a legitimacy-centered framework demonstrating that implementation gaps represent strategic incompleteness preserving stakeholder consultation opportunities. Our findings challenge efficiency-focused AI strategy literature, showing that mission-driven organizations must privilege accountability over optimization when conflicts emerge. This research provides first systematic evidence of how bureaucratic institutions navigate algorithmic transformation while protecting democratic legitimacy.