Artificial Intelligence in Agribusiness Transformation: Advancing Precision Farming under the United Nations Sustainable Development

Abstract

Precision has an increasing dependency on dependable automated decision-making systems and trustworthy sensor data; however, resource inefficiency, cyber manipulation, and unequal access limit the benefits in the agribusiness sector. This study looks at a SHA-256 blockchain security integrity ledger called the IoT (Internet of Things) along with real time Digital Twin applications and small hold farming adoption in order to achieve the sustainable goals 2,6,9, and 12 of the United Nations. A sequential method which integrates these designs was used. The orthodox pipeline was compared with a proposed pipeline using cryptographic verification. Across 3,000 test records, end-to-end operational decision accuracy increased from 68.83% to 91.80%, while the false-positive irrigation rate fell from 12.48% to 1.82%. Cryptographic verification detected 100.00% of the simulated malicious alterations with a 0.00% integrity false-positive rate, and the anomaly-detection stage achieved 81.35% recall and an 80.35% F1 score. A paired comparison of decision correctness was statistically significant (exact McNemar p < 0.001). Qualitative conceptual formation further indicates that Digital Twin-AI systems can strengthen real-time resource decisions, whereas smallholder adoption remains held back by digital literacy, infrastructure, affordability, local relevance, and governance. The findings support an integrating technology and governance approaches in which secure data pipelines are combined with participatory and context sensitive adoption strategies to improve precision farming sustainably and inclusively.