Marine biodiversity is essential for the health and sustainability of ocean ecosystems, yet increasing human activities and climate change are causing significant declines. Traditional taxonomic approaches, heavily reliant on taxonomic keys, are often hampered by limited expertise, geographic constraints, and incomplete resources. This study leverages the transformative potential of artificial intelligence (AI) and machine learning (ML) to address these challenges, employing advanced image, video, and acoustic analysis alongside generative AI-driven tools like TAXObot for species identification. TAXObot, tailored for marine taxonomy, achieved an accuracy of 98.33%, significantly outperforming leading Large Language Models (LLMs) in taxonomic query resolution. Cutting-edge models based on visuals and acoustic classification demonstrated remarkable efficiency in species detection and classification, while structured datasets and domain-specific preprocessing proved pivotal in ensuring precision. Despite challenges such as dataset imbalance and novel species detection, emerging AI offer promising solutions. This study establishes a foundation for integrating AI/ML into marine biodiversity research, advocating for scalable, adaptive tools that enhance our understanding of ocean ecosystems and bolster global conservation efforts.