A Systematic Literature Review of Deepfake Literacy, Societal Impacts, and AI-Driven Prevention Strategies for Platform-Specific Vulnerabilities

Deepfake technology has rapidly evolved from a niche novelty into a significant threat to digital security, media integrity, and societal trust. This systematic review aims to analyze the progression of deepfake technology, its increasing societal impact, and how generative AI models can be leveraged for developing personalized, platform-specific prevention strategies. The review begins by exploring the foundational work on deepfakes, focusing on the technological advancements of Generative Adversarial Networks (GANs) that make the creation of hyper-realistic manipulated content possible. Then it examines the social implications of deepfakes, including their role in misinformation campaigns, political manipulation, and privacy violations. As deepfakes become more prevalent across social media platforms like Facebook, Instagram, and YouTube, the challenges in detecting and mitigating their impact are increasing. The review highlights the limitations of current detection tools and the urgent need for user education and technological interventions. Generative AI models such as ChatGPT, Google Bard, and Meta’s LLaMA are analyzed for their potential to provide tailored deepfake prevention strategies that enhance user awareness and digital literacy. The study compares the ability of these models to offer personalized recommendations, focusing on the superior adaptability and user-friendliness of ChatGPT. By addressing the gaps in deepfake literacy and platform-specific vulnerabilities, this review aims to present a comprehensive framework for mitigating the risks associated with deepfakes and empowering users with AI-driven solutions.