- VA has requested industry feedback on third-party enterprise AI services
- Industry responses to the notice are due Oct. 7
- The 2026 Healthcare Summit on Dec. 3 will explore AI governance and more
The Department of Veterans Affairs has issued a request for information seeking input on third-party services to support the deployment, adoption and governance of an enterprise artificial intelligence product across its workforce.
As federal health agencies expand the use of AI tools among clinicians, benefits staff and other personnel, questions about training, governance and measurable outcomes are taking on greater weight. The Potomac Officers Club’s 2026 Healthcare Summit on Dec. 3 will bring government and industry leaders together to discuss AI governance and policy, balancing cost, quality and access to care, and military health readiness. Book your spot now to hear from officials guiding federal health modernization.
In a SAM.gov notice posted Tuesday, VA said the notice is the last installment in a three-part series of industry queries supporting its enterprise AI acquisition strategy. The department expects to release a solicitation for the services in early fiscal year 2027, possibly as soon as October.
Responses are due Oct. 7.
What Problem Areas Will the Third-Party Services Address?
The anticipated contract covers support services that would operate alongside an enterprise AI product VA plans to obtain through a separate procurement, with the selected contractor coordinating with the department and the product’s supplier. According to the notice, the third-party contractor’s role would center on running software within VA’s environment, supporting and communicating with users, and identifying problems to solve through tools built on top of the product. A draft performance work statement attached to the notice outlines a 12-month base period and two 12-month option periods.
The notice identifies 10 problem areas for the contractor to address:
- IT operations: The contractor would work with VA teams on networking, identity, endpoint, help desk, security and sustainment functions.
- Discovery and adoption: Customer success leads, up to 50 in total, would help users across VA business units find and use the product’s features.
- Role-based virtual training: Problem-based instruction would be tailored to job functions such as clinical care, benefits processing and acquisition.
- Champion training: Selected users would receive advanced instruction to prepare them to train colleagues.
- Data connectivity: The work includes integrating data connections and setting an agencywide standard for maintaining them.
- VA-built AI assistants and agents: A small development team would build custom tools, with capacity growing as demand rises.
- Responsible AI adoption: The contractor would help keep product use aligned with federal laws, regulations and VA guidance.
- User feedback loop: User input would be gathered to inform the product roadmap.
- Financial monitoring: The effort includes recurring return-on-investment tracking and financial operations management.
- Migration support: VA would receive help moving users to a different product if it decides to switch.
The department also invited industry to propose problem areas beyond those listed.
What Information Is VA Seeking From Industry?
VA is asking respondents to address the following:
- Capabilities and experience: Companies should describe their ability to perform the work outlined in the draft PWS, including prior support for enterprise or generative AI environments in federal or other regulated, high-security settings, along with relevant certifications, partnerships and provider escalation capabilities.
- Subcontracting: The department wants to know what portion of the work companies would expect to assign to subcontractors.
- PWS feedback: Vendors are asked whether the draft offers enough detail to respond to a request for proposals and price the work on a firm-fixed-price basis, and to identify requirements, assumptions, risks or ambiguities that could affect competition, pricing or staffing.
- Commercial practices: For problem areas 2 through 10, VA is seeking examples of typical services and deliverables in government and industry and how those offerings are commonly priced.
- Technical challenge ideas: VA plans to include a technical challenge in its evaluation and wants suggestions for demonstration-based approaches.
- Payment milestones: Respondents should recommend how to structure firm-fixed-price payment milestones for each period of performance and whether those milestones should be based on time, performance, deliverables or a combination.
- Rough order of magnitude pricing: The department is requesting estimates for PWS tasks 5.1 through 5.5, broken down by labor type, mix and hours, along with the main factors that would drive costs.
- Additional considerations: Vendors may raise other issues the government should weigh when evaluating and contracting for the support.
What Other AI-Related Efforts Has VA Pursued?
The latest RFI follows other VA initiatives involving AI. In October 2025, the department published an AI adoption strategy centered on building data infrastructure, reshaping core workflows and developing an AI-ready workforce, and sought industry feedback on an IT Service Management Platform that integrates AI and offers low-code tools. A November 2025 RFI on the VA Enterprise Cloud sought recommendations on its future design and acquisition strategy, including AI integration, and VA issued another RFI in June on a potential enterprise cloud brokerage service.
Addressing expectations for industry, Zack Schwartz, principal deputy assistant secretary in VA’s Office of Information and Technology, said continued IT partnerships with the department will depend heavily on contractors’ ability to support system modernization, AI integration and shifting mission demands.
