VA officials and lawmakers debate AI’s role in disability claims processing amid staffing concerns

During a House Veterans’ Affairs Technology Modernization Subcommittee hearing, Department of Veterans Affairs officials stated that AI tools used to expedite disability claims processing are designed to support human reviewers rather than replace them. Robert Orifici, acting deputy chief information officer at VA’s Office of Information and Technology, testified that every disability claim is decided by a trained VA employee and that AI tools are intended to augment productivity on high-volume administrative tasks.

Derek Herbert, acting chief production officer at VA’s Veterans Benefits Administration, added that the agency focuses on automating routine tasks to free staff for higher-value work, emphasizing that AI is a productivity tool rather than a replacement for human reviewers.

Rep. Nikki Budzinski, D-Ill., challenged the effectiveness of VA’s AI deployments, citing employee reports that automation and AI pilots have produced incorrect information, which she argued slows processing and undermines accuracy. She also expressed concern that technology alone cannot address staffing shortfalls, noting VA has lost 1,100 claims examiners in the current fiscal year and 2,700 examiners and 120 IT specialists since January 2025.

Rep. Maggie Goodlander, D-N.H., echoed concerns about staffing levels, pressing VA officials for specific examiner counts and confirming declines since January 2025. VA officials acknowledged staffing gaps but did not provide exact numbers, stating they were hiring and would supply the data to lawmakers.

Sterling Thomas, chief scientist at the Government Accountability Office, recommended that federal agencies develop a workforce with the requisite skills to deploy AI responsibly. He suggested establishing a federal digital services academy and following GAO’s AI framework, which calls for documenting training data, defining system specifications, continuously monitoring performance, and conducting regular assessments to ensure responsible AI deployment.