NIST is applying AI to its quantum-based measurement tools for mass, voltage, and resistance. Digital twins will control these instruments using physics-based AI models. The models improve accuracy and robustness while making calibrations faster and simpler. Researchers can test and refine procedures virtually before changing physical systems. The work covers the Kibble balance for mass, Josephson voltage standards, and quantum Hall effect for resistance. AI models incorporate physical constraints to ensure scientifically grounded results. The project aims to make advanced measurement capabilities available to external partners. A prior project already proved AI could improve Kibble balance mass measurements. This effort is part of NIST’s SERI program and is ongoing. The team uses offline models and physical constraints to address safety concerns.