Artificial intelligence (AI) uses machine-learning algorithms and software that mimic human cognition in the analysis, presentation, and comprehension of complex data. AI is increasingly being used in many domains of human life. Physicist and author Stephen Hawking mentioned, “Perhaps we should all stop for a moment and focus not only on making our AI better and more successful but also on the benefit of humanity.” An increasing number of studies relating the successful role of AI in diagnostic medical imaging echo the above thought of Stephen Hawking. Within ophthalmology, the role of AI is increasingly being realized for the screening of potentially blinding diseases, particularly diabetic retinopathy (DR).DR is the leading cause of blindness among the working-age population of both the developing and the developed countries.[1] The management of DR involves multiple visits to the eye hospital and poses a significant economic burden on the patient. However, early detection, proper management of the systemic and ophthalmic conditions, and good compliance to treatment can limit the vision loss due to DR. Screening of the diabetic population is crucial in the early detection of DR.Numerous studies have demonstrated the role of AI in the screening and detection of DR from color fundus photos with high sensitivity and specificity. The landmark study by Abràmoff et al.[2] demonstrated the feasibility of AI-enabled screening of DR in primary care settings. A recent article in the Indian Journal of Ophthalmology demonstrated the robustness of offline AI algorithms in a smartphone-based fundus camera for community-based DR screening and is a major advance in the field of screening of DR.[3] Of note, the same group of authors validated the use of a similar offline AI algorithm to provide real-time analysis of the images captured on a smartphone in a pilot study of 231 patients in a previous study.[4] Smartphone-based fundus camera has better accessibility, portability, and cost-effectiveness than the conventional tabletop fundus cameras. This makes the smartphone-based fundus camera a preferred tool for the community screening of DR. The retinal images were captured by a trained optometrist and were subjected to the proprietary offline automated analysis on a smartphone to detect referable diabetic retinopathy (RDR) taken as grades of moderate nonproliferative diabetic retinopathy (NPDR) or higher severity of NPDR and proliferative diabetic retinopathy (PDR).[3] The capturing of the images by the trained optometrists addresses the scarcity of trained vitreoretinal surgeons especially in developing countries like India. The offline automated smartphone-based analysis of the images makes the method adaptable in the community, including the remote and rural locations where round-the-clock working internet connection may not be available. The inclusion of moderate NPDR or higher severity of NPDR and PDR in the RDR group is similar to another landmark study by Gulshan et al.[5] in which the authors retrospectively studied 128,175 retinal images from EYEPACS-1 and MESSIDOR-2 data sets and demonstrated high sensitivity and specificity for RDR.The authors in the current study have demonstrated a sensitivity and specificity of diagnosing RDR as 100% (95% CI: 94.72%–100.00%) and 89.55% (95%CI: 87.76%–91.16%), respectively, which is quite impressive.[3] The authors have validated the algorithm in a prospective study, which not only supports but also is statistically preferable to many prior retrospective studies to detect DR in color fundus photographs with high sensitivity and specificity. This algorithm will help in the detection of DR at the patients’ doorsteps, and visiting an eye care system for the screening of DR may not be mandatory for the patients.The only aspect where the study could have been stronger is the inclusion of diabetic patients for screening from multiple geographical areas and/or from different ethnicities. Currently, the authors included patients visiting 47 different dispensaries administered by the Municipal Corporation of Greater Mumbai, India.[3]There are some future perspectives in the field of smartphone imaging integrated with offline AI. First, similar to DR, algorithms can be developed for classifying diabetic macular edema (DME) as referrable and nonreferrable. Although there are some reports of high diagnostic accuracy of detecting DME on smartphone-based screening of DME with offline AI, multicentric, multiethnic, and prospective studies will be more useful.[6] Second, like many other smartphone applications, the development of an offline AI-enabled application and integration with the smartphone will help the patients diagnose them as RDR and referrable DME (RDME). This will further simplify the procedure of early detection of RDR and RDME. Third, it will be better to study smartphone imaging integrated with offline AI in the eyes without mydriasis. In the current study, the eyes of the patients were dilated before the acquisition of retinal images. Fourth, it will be better to conduct prospective studies on smartphone imaging integrated with offline AI in the detection of diseases such as glaucoma and age-related macular degeneration (ARMD). Screening and early detection of potentially blinding diseases such as DR, glaucoma, and ARMD will prevent vision loss, and smartphone imaging integrated with offline AI holds a lot of promise in this regard.