How can videos be used and maximized in global ocean research? And how can ocean videos be better accessed in classrooms? Can these data give us a better look at evolving underwater environments over time? Labs across NOAA collect ocean video using an array of technologies, methods, and formats to support Earth research objectives. Videos are made scientifically useful through textual descriptors, called annotations. Some labs annotate video manually, while others train ML algorithms to identify sealife and geological features of interest. Due to the complexity and variety of practice, archiving these critical video data and their associated annotations has been time consuming. A metaanalysis of current NOAA video practices is needed to properly expand the current archival pipeline and public access through NOAA's Video Portal (https://www.ncei.noaa.gov/access/ocean-exploration/video/). In response, CIRES researchers are collaborating with NOAA's National Centers for Environmental Information (NCEI) and NOAA's Center for Artificial Intelligence (NCAI) to 1) analyze all video data collections and AI/ML workflows across NOAA and 2) recommend a standardized approach for video collection, analysis, archival, and stewardship. This ongoing work will help document NOAA-wide video collections and AI applications to maximize benefits for users at NOAA and beyond.