AI-Driven Detection and Mitigation of Disinformation Threats to U.S. National Security: A Systematic Review

The proliferation of artificial intelligence (AI) tools has fundamentally transformed the nature and scale of disinformation threats confronting the United States. State-sponsored actors, principally Russia, China and Iran, deployed AI-enhanced influence operations targeting U.S. democratic institutions throughout the 2022 midterms and the 2024 presidential election, using synthetic audio, deepfake video and algorithmically generated text to polarize public opinion and undermine confidence in electoral processes. The U.S. Government Accountability Office identified foreign disinformation as one of the most pressing national security challenges facing the Departments of State, Homeland Security and Defense. Concurrently, researchers and intelligence agencies face an arms-race dynamic: generative AI simultaneously enables disinformation at scale and furnishes the computational tools required to detect it. This systematic review synthesizes the peer-reviewed and governmental literature published between 2019 and 2024 on AI-driven disinformation detection and mitigation. The review examines four detection technology classes: natural language processing (NLP) classifiers, deepfake forensics, network propagation analysis and content provenance systems, alongside the U.S. policy and institutional frameworks designed to counter these threats. Key findings indicate that transformer-based NLP models, particularly RoBERTa and BERT-based architectures, achieve classification accuracies of 88-95 percent on benchmark disinformation datasets, and multimodal detection systems that integrate text and visual features outperform unimodal approaches by 6-14 percentage points. However, significant gaps exist in real-time detection scalability, which includes generalization across linguistic and cultural contexts and coordination among federal agencies. This review concludes with a taxonomy of open research challenges and a set of evidence-based policy recommendations for strengthening the U.S. national security response to AI-driven disinformation.