Ecological Risk Assessments (ERAs) are widely adopted in Australian fisheries to evaluate the potential impacts of fishing activities on interacting species and determine their risk level based on biological characteristics, population status, and ecological attributes. Maintaining these attributes is critical for ensuring assessments reflect current scientific knowledge; however, traditional manual updates can be time-consuming and resource-intensive. Agentic Artificial Intelligence (AAI) offers a promising approach to improving the efficiency of attribute updates by rapidly searching, identifying, extracting, and synthesising information from online literature, databases, and technical reports. This can support more timely incorporation of new attributes into ERA processes and assist in detecting knowledge gaps. Despite these benefits, challenges remain regarding data quality, transparency, uncertainty, and the need for expert validation of information obtained by AAI. Consequently, AAI is viewed as a decision-support tool rather than a replacement for ecological expertise. This paper evaluates the application of AAI for updating ecological attributes used in fisheries ERAs and compares AAI-updated information with attributes currently incorporated in assessment frameworks. The findings highlight the potential of AAI to enhance the efficiency, consistency, and adaptability of future ERAs.