A review of responsible use of AI in remote sensing and cartography

The integration of Artificial Intelligence (AI) into remote sensing (RS) offers profound capabilities in data analysis, interpretation, and knowledge discovery. While AI applications offer immense potential for addressing global challenges such as environmental monitoring, disaster management, and urban planning, their rapid increase necessitates a critical examination of responsible development and deployment practices. This review article synthesizes existing scientific literature to explore the multifaceted dimensions of responsible AI in remote sensing, focusing on six core aspects: (1) fairness and bias, (2) privacy and data protection, (3) ethical and societal impacts, (4) transparency and explainability, (5) trustworthiness and robustness, and (6) accountability and governance. The analysis identifies key challenges, including the perpetuation of existing inequalities through biased data, as biased AI models trained on satellite imagery could lead to inequitable resource allocation, the erosion of individual privacy by high-resolution surveillance, and the complexities of ensuring accountability in opaque AI systems. Furthermore, it discusses current mitigation strategies, such as advanced bias detection techniques, privacy-preserving machine learning paradigms like federated learning, and robust regulatory frameworks. Ultimately, the challenge is not just to build more advanced AI but to cultivate a global ethical framework that ensures these powerful tools serve, rather than control, the planet and its people. This work uniquely contextualizes responsible AI principles within remote sensing and cartography, offering a domain-specific synthesis and actionable research directions to bridge the gap between high-level ethical frameworks and practical geospatial applications.