Veterans Affairs’ Office of Inspector General uses Microsoft Copilot to guide reviews of care sites
The regulatory agency responsible for monitoring the U.S. Department of Veterans Affairs has been using an AI chatbot as part of its investigations, a media report revealed.
The VA’s Office of Inspector General (VA-OIG) has been using Microsoft Copilot to help it collect data on patient care facilities run by the department, in an effort to ensure they are in compliance with federal law and delivering quality care.
According to NextGov, the agency has been using the artificial intelligence application to do more than pull information, with the agency also relying on the model to read and analyze documents, including financial and medical records.
The use of the AI is a bit of an open secret. As the outlet pointed out, the VA-OIG has included a disclosure statement in all of its inspection reports since March, when the agency stated that its “inspection team used peer-reviewed standardized, structured, and evaluated prompts in Microsoft Copilot Chat to review inspection data such as interview transcripts, documents, questionnaire responses and physical observation.”
The agency goes on to add that it takes “full responsibility for the content” of its reports, adding that a human investigator is tasked with confirming their accuracy.
Use of the AI is linked to audits of VA hospitals, clinics, providers and mental health services, by the agency’s own admission.
In a statement to NextGov, the VA-OIG said the AI chatbot is used to improve efficiency, promising that sensitive patient data remains secure: “[Data] entered into this tool is not shared with the Department, nor does it leave the VA-OIG secure environment,” a spokesperson is quoted as saying.
Citing a source at the agency, reporters said Microsoft Copilot has been in use since 2024, though it isn't clear when regulators began using it for audit reports.
As NextGov mentioned, the policy aligns with an April 2025 memo from the Office of Management and Budget that attempted to set parameters for when federal agencies can use AI, which does not include instances where the models are used for actionable decision making.
The VA-OIG stated that it does not use the AI chat for any of those high-impact functions, such as deciding when to sanction a veteran healthcare clinic.
For more, read the full story at the link below.
Chad is an award-winning writer and editor with over 15 years of experience working in media. He has a decade-long professional background in healthcare, working as a writer and in public relations.