AI-enabled identity verification frameworks for safer and more inclusive U.S. financial services

AI-enabled identity verification has become essential to fraud protection and customer onboarding in U.S. financial institutions, but its rapid adoption has generated concerns about exclusion and algorithmic bias. Traditional verification regimes based on documentation identities, credit histories and steady residency continue to exclude people with thin files, insecure housing or non-standard legal status, while developing AI systems risk perpetuating these injustices in digital form. This study investigates whether AI-based identity verification frameworks improve security while also increasing financial inclusion, or if they mainly automate previous patterns of exclusion. The study develops a sociotechnical analytical framework by conducting an interdisciplinary literature synthesis that includes peer-reviewed scholarship from 2015 to 2026, as well as regulatory reports, industry white papers and documented implementation cases. A comparative investigation of traditional banks, fintech firms and community-based financial institutions finds substantial demographic performance discrepancies in biometric and alternative data systems, which are influenced by institutional incentives and regulatory uncertainty. The results indicated that no single verification modality delivers uniform equity and that institutional emphasis on fraud minimization frequently outweighs inclusion goals. The study concludes by proposing a contextual, multidimensional verification framework based on procedural fairness, demographic performance monitoring, and regulatory safe harbors, which argues that equitable AI verification requires deliberate governance and long-term collaboration among regulators, technologists and financial institutions. Keywords: AI, Identity Verification, Financial Inclusion, Algorithmic Bias, Alternative Data, Biometric Authentication.