In the current era of digital technology, the creation of photos produced in an activity is experiencing a significant amount of "explosion". These photos are often stored without adequate metadata, allowing a great deal of information and events to fade over time. In this case, information managers and archivists have a big task dealing with still image archives that must be identified and analysed immediately. Currently there has been research on how Artificial Intelligence (AI) can be used to help collect metadata and descriptions of still image archives. This paper will describe how still image archive identification activities are currently carried out at the Indonesian National Archives manually and provide an overview of how this activity can be assisted with the use of AI technology. It is hoped that the results obtained from this paper can serve as a basis for further research regarding the use of AI for archival activities in general, especially for Indonesian audio visual and still image archives such as photographs and other works of art, and can be a solution for creating more detailed archival descriptions and can later help improve findability in still image archives searches.