Huanglongbing (HLB), commonly known as Asian citrus greening, is a devastating citrus disease associated with ‘Candidatus Liberibacter asiaticus’ (CLas), a pathogen transmitted by Diaphorina citri. This psyllid species is of Asian origin, and since the mid-2010s has been recorded in Kenya, Ethiopia, Nigeria, Tanzania, Benin, and Ghana, while CLas is present in Ethiopia and Kenya. HLB is also vectored by Trioza erytreae, a psyllid species indigenous to sub-Saharan Africa. Monitoring, early detection, and timely identification are needed to manage the vectors, and consequently HLB. Currently, psyllid identification is expensive due to the use of yellow sticky traps and the need for specialists for time-consuming screening of traps and species identification. Therefore, we aimed to develop a laboratory-based, artificial intelligence (A.I.)-driven diagnostic tool for citrus psyllid pests, while accounting for native non-pest Diaphorina species, on yellow sticky traps. We deployed 544 traps across multiple citrus growing areas within South Africa, Reunion Island, and Mauritius, between December 2021 and September 2023. The traps were collected and inspected for D. citri, native South African Diaphorina species, and T. erytreae. Traps with any of the target psyllid species were photographed using a digital RGB camera, which were annotated highlighting those species. The annotated images were processed and augmented to develop and train five YOLOv8 models using a 7:2:1 training, validation, and testing ratio. Models YOLOv8s and YOLOv8m show promise for rapidly identifying psyllids to permit faster implementation of control methods. Both models achieved a mean average precision of 0.85, taking 6.7 seconds and 12.7 seconds, respectively, to process a two-sided trap. These models showed potential for the initial screening of yellow sticky traps to identify those needing verification of psyllid presence.