AI-Powered Influencer Video Auto-Posting Platform with Smart Captioning, Hashtag Generation, and Fake Video Detection: A Review

Abstract

The rapid growth of user-generated multimedia content on the web has amplified the need for automated systems to generate captions, personalise hashtags, and detect deepfakes. Not only do these technologies increase user interaction, but they also play a significant role in protecting trust and authenticity online. Previous surveys have widely studied these areas separately. Captioning studies of image reviews have progressed from CNN-RNN models to attention and transformer models, and have mostly optimised for fluency and coherence. Hashtag recommendation research focuses on personalisation approaches, often using user behaviour modelling to enhance content exposure. Likewise, deepfake detection surveys point to multimodal fusion and adversarial learning methods to enhance detection strengths. Though these contributions are noteworthy, they tend to be domain-specific, failing to account for cross-task dependencies and real-world deployment issues. Additionally, repeated constraints, such as English-language bias in datasets, poor handling of low-resource settings, and inadequate attention to trustworthiness and ethics, limit their generalizability. This review addresses these limitations by bringing research in captioning, hashtag recommendation, and deepfake detection under a single umbrella of social media automation. By systematically connecting generation, personalisation, and verification, it delineates common challenges such as dataset bias, real-world generalisation, and scalability, while charting bleeding-edge solutions that aim to span these gaps. In contrast to previous disjointed surveys, this review focuses on the interaction between technical design and platform credibility, offering a unifying perspective. Finally, it lays the groundwork for future research that progresses not just algorithmic performance but also the ethical, reliable incorporation of AI-based multimedia systems into digital ecosystems.