Data-based technologies are increasingly embedded in supply chain decision-making, where they transform distributed operational, environmental, and organisational data into forecasts, summaries, recommendations, and decision-support outputs. While these technologies can improve visibility, responsiveness, sustainability performance, and resilience, they may also amplify risks when outputs are inaccurate, opaque, weakly validated, or used without clear accountability. This paper develops a governance-oriented framework for data-based technologies in sustainable and resilient supply chain decision-making, with large language models used as an illustrative case. The framework is structured around four governance checkpoints: application context, regulatory sensitivity, model validation, and impact and feedback monitoring. The paper contributes a conceptual basis for integrating governance into data-supported decision processes and highlights the importance of human oversight, validation, auditability, and feedback learning in preventing risk escalation.