HUMINT：HUMAN INTELLIGENCE AN AUTHORITATIVE LANDSCAPE REPORT ON EMERGING TECHNOLOGIES

Generative artificial intelligence, deepfakes, and synthetic identities are reshaping the information acquisition and verification environment in human intelligence (HUMINT). Yet important knowledge gaps remain: mature methods are often undervalued, pseudo-frontier technologies are overstated, source reliability is conflated with information credibility, and the boundaries of AI responsibility remain unclear. This study applies an evidence-native open-source synthesis framework that integrates international norms, government standards, systematic reviews, meta-analyses, empirical research, and authoritative R&D programs. Source grading, structured analysis, an eight-dimension evaluation framework, and a maturity–risk matrix are used to identify technical effectiveness, scope conditions, and governance requirements. The findings indicate that, relative to accusatorial and pressure-based strategies, rapport-based, noncoercive information-gathering interviews are positively associated with the production of verifiable information, in part because they reduce defensiveness, suggestibility, and memory contamination. AI can augment transcription, cross-language retrieval, entity resolution, identity proofing, and training; however, independent truthfulness adjudication, machine-learning lie detection, and automated emotion recognition still lack adequate external validity. The report therefore reconceptualizes HUMINT as a sociotechnical system in which professional judgment, verifiable evidence chains, and constrained AI operate together. It proposes a capability stack, deployment matrix, and quality-gate framework to support institutional training, technology procurement, audit oversight, and the protection of rights.