The Department of Veterans Affairs Office of Inspector General has been using Microsoft Copilot Chat since 2024 to review documents, interview transcripts, and other data collected during audits of VA healthcare facilities. The tool is deployed within the Office of Healthcare Inspections, which evaluates whether veterans receive proper medical care and mental health services. Investigators do not use the AI to make final decisions or write reports, but to flag patterns and potential issues across large volumes of inspection data.
Since March, OIG hotline and inspection reports have carried a standard disclosure confirming the use of Copilot Chat. The boilerplate language states that "during the preparation of this report, the inspection team used peer-reviewed standardized, structured, and evaluated prompts in Copilot Chat (Microsoft) to review inspection data such as interview transcripts, documents, questionnaire responses, and physical observations." It adds that investigators "confirmed fidelity of the generated output to the source material, edited the report, and took full responsibility for the content of the publication."
How the watchdog uses the tool
Former VA OIG Inspector General Michael Missal, who was fired in January 2025 as part of a mass purge of federal agency IGs, said his office tested and purchased Copilot Chat with the goal of making inspections faster and more consistent. Site audits of VA medical facilities often examine the same issues repeatedly, so standardized prompts helped ensure teams did not overlook key areas.
"It's a time-saving way [to conduct inspections], and it's also quality assurance as well because this way everybody's operating with the same prompts, and it just ensures you're not missing something," Missal said. "What this does is it prompts you: 'What about this? What about this? You get this. What about that?' You know, [it] just takes you through a whole menu of things."
The watchdog's spokesperson said data entered into Copilot Chat "is not shared with the Department, nor does it leave the VA OIG secure environment." All AI-generated output requires human review before any information appears in a published report. Investigators verify every finding against original source material.
Not classified as high-impact AI
An April 2025 memo from the Office of Management and Budget directed agencies to manage high-impact AI uses - cases where AI output "serves as a principal basis for decisions or actions that have a legal, material, binding, or significant effect on rights or safety." The OIG determined that Copilot Chat does not meet that threshold because inspection teams "do not use AI as the principal basis for decision-making or actions."
The watchdog's 2025 AI use case inventory lists 20 examples of emerging capabilities, with one entry specifically identifying Copilot Chat for use in word processing tools. A separate OIG report from June warned that Veterans Health Administration personnel were using Copilot Chat for clinical purposes, even though the agency had not designated it as a high-impact AI use case.
Missal said he had planned to expand Copilot Chat's use to the Office of Audits and Evaluations and the Office of Investigations before his removal. For now, the tool remains concentrated in healthcare inspections, where structured, repeatable review work aligns with what the Microsoft AI Courses ecosystem supports. The OIG's AI strategy document states the office uses AI "in meaningful ways that enhance the office's mission of conducting independent oversight that improves the health and welfare of veterans and their families, caregivers, and survivors."
Why this matters for healthcare professionals
For clinicians and administrators working in large health systems, the VA OIG's approach offers a practical model for AI adoption: use standardized prompts to surface patterns in audit data, but keep humans firmly in control of final judgments and written reports. The disclosure language appearing in public inspection documents also sets a transparency benchmark. Any healthcare organization adopting AI for Healthcare quality reviews can learn from this structure - standardize the prompts, lock down the data environment, and make the AI's role explicit in published findings. The line the OIG draws between efficiency support and high-impact decision-making is one that hospital compliance teams and quality officers will need to navigate as these tools spread.
Your membership also unlocks:
