VA AI Contract Strategy Raises Veteran Care Stakes

The Department of Veterans Affairs is signaling that healthcare technology contractors will be judged less by incumbency and more by their ability to support modernization, artificial intelligence, and measurable operational outcomes. That message may sound like a procurement shift. In practice, it is a care delivery issue.

The Department of Veterans Affairs operates one of the nation’s largest and most complex healthcare environments. Its technology decisions affect clinical documentation, appointment access, pharmacy operations, benefits administration, cybersecurity, revenue processes, and coordination between military and veteran care. The VA Office of Information and Technology describes its mission as delivering a seamless, unified Veteran experience through technology. (U.S. Department of Veterans Affairs)

A harder line on vendor performance is reasonable. Healthcare technology contracts can outlast leadership teams, strategy changes, and the tools they were initially designed to support. A contractor that met prior requirements may not be equipped to support a new AI-enabled workflow, cybersecurity expectation, interoperability standard, or clinical modernization effort. The risk is not contract turnover. The risk is allowing long-standing vendor relationships to become a substitute for progress.

Procurement Is Becoming a Care Delivery Lever

Federal healthcare procurement often appears distant from clinical work, but contract decisions shape what clinicians and patients experience. A vendor responsible for infrastructure, support services, identity management, EHR integration, data exchange, or workflow automation can affect whether a care team gains access to reliable information at the moment it is needed.

That makes VA’s message to contractors more consequential than a warning about competitive rebidding. It is an attempt to reposition technology acquisition around current mission needs rather than historic relationships. The agency’s 2026 Advanced Planning Brief to Industry specifically centered VA acquisition and IT transformation priorities, with senior Office of Information and Technology leaders participating in the event. (DigitalVA)

This approach could improve accountability if performance measures are tied to meaningful outcomes. Faster delivery of a tool is not enough. A contractor should be able to show whether its work reduces clinician burden, improves system reliability, strengthens cybersecurity, shortens service delays, supports data exchange, or improves access for veterans.

AI should fit inside that same standard. A vendor should not receive credit merely for adding an AI feature to an existing product. The relevant question is whether the feature improves a defined workflow without adding safety risk, privacy exposure, unreliable output, or new administrative work.

AI Requirements Must Be Linked to Measurable Outcomes

The rapid expansion of AI has created pressure across healthcare to demonstrate innovation quickly. Federal agencies face the same pressure. The White House’s current federal AI policy encourages agencies to accelerate AI use while maintaining governance and public trust. (AI.gov)

That balance is essential in the VA environment. AI may help with help desk support, data analysis, scheduling, clinical documentation, cybersecurity monitoring, call center operations, benefits navigation, and workflow automation. Each use case carries a different risk profile.

An AI system used for internal administrative triage may require a different level of oversight than one influencing clinical decision support, patient communication, or eligibility-related workflows. Contractors should therefore be required to define intended use, data sources, access permissions, testing methods, human review requirements, error handling, and performance monitoring before deployment.

The National Institute of Standards and Technology provides a useful framework through its AI Risk Management Framework, which emphasizes governance, risk mapping, measurement, and continuous management. (AI.gov) The framework is particularly relevant for VA because AI tools may operate across sensitive environments containing veteran health records, benefits data, identity information, and operational systems.

The strongest procurement model will require contractors to prove that AI is improving service delivery rather than creating a new layer of technical complexity. AI must be treated as a capability with measurable obligations, not as a procurement category that automatically signals modernization.

EHR Modernization Cannot Be Treated as a Separate Track

VA’s AI ambitions also intersect with the Federal Electronic Health Record modernization effort. The agency’s current deployment schedule states that the Federal EHR is planned for deployment across 164 VA medical centers and associated clinics beginning in 2026, while 10 medical centers are already live on the platform. (DigitalVA)

That scale makes the connection between EHR modernization and contracting unavoidable. AI cannot operate safely on top of incomplete, inconsistent, or unreliable clinical data. A decision support tool, documentation assistant, workflow automation product, or patient engagement platform depends on accurate information moving through the EHR and related systems.

The Government Accountability Office has continued to identify VA healthcare and EHR modernization as a high-risk area, citing the scale of the remaining implementation work and the need for stronger oversight. (Government Accountability Office Files) In a more recent assessment, GAO recommended that VA address unresolved critical and high-severity testing findings before deploying the new EHR at additional locations. (GAO)

Those findings should influence how VA evaluates contractors. AI capability cannot compensate for unresolved EHR safety, usability, or reliability problems. A sophisticated automation layer placed on unstable clinical infrastructure may make existing workflow failures harder to detect and harder to correct.

The proper sequence is clear. EHR reliability, data quality, cybersecurity, and workflow usability should be established before AI is expected to deliver large-scale value. Contractors that support those foundations may be more strategically important than those offering the most visible AI features.

Contract Flexibility Needs Stronger Governance

Reassessing contracts as requirements change can improve responsiveness. Healthcare technology needs evolve quickly, particularly when new cyber threats, clinical demands, regulatory expectations, and interoperability standards emerge. A multi-year agreement should not lock VA into a solution that no longer fits the mission.

At the same time, flexible contracting needs disciplined governance. Frequent changes in direction can create implementation fatigue, fragment accountability, and leave frontline staff adapting to overlapping tools. A vendor transition may improve long-term performance while still creating short-term operational risk if data migration, training, workflow redesign, and continuity planning are weak.

VA should establish clear decision criteria for evaluating whether a contractor remains aligned with agency needs. Those criteria should include system availability, cybersecurity posture, interoperability performance, clinician usability, support responsiveness, data quality, cost transparency, implementation milestones, and veteran-facing outcomes.

A contractor that loses a contract because priorities changed should receive a clear explanation of the standards applied. That transparency is necessary for a competitive market and for federal accountability. It also protects VA from replacing one incumbent problem with another vendor that offers stronger marketing than operational readiness.

Cybersecurity Must Be Part of Vendor Performance

AI adoption and modernization expand the cyber risk surface. More automation can mean more integrations, more identities, more APIs, more data exchange, and more third-party access. Each of those elements requires governance.

VA technology contractors should be evaluated on how well they protect veteran data, monitor access, manage privileged accounts, secure integrations, test recovery, and respond to incidents. Cybersecurity cannot remain a separate compliance appendix attached to modernization work. It must be a central measure of whether modernization is safe.

The VA relies on technology systems to store and process sensitive veteran information, including medical records. GAO has stressed that VA’s IT systems are essential to delivering services and maintaining sensitive data, making modernization and security inseparable. (Government Accountability Office Files)

A contractor that adds AI functionality without strengthening identity governance, data segmentation, logging, and recovery planning is not advancing the mission. It is increasing exposure. Procurement standards should reflect that reality.

The Standard Is Veteran Outcomes

The most constructive interpretation of VA’s tougher stance is that contractors are being asked to compete on outcomes. That is the right direction. Technology vendors should not be insulated by tenure when veterans and clinicians need systems that work better than they did a year ago.

But VA should resist equating speed with success. Rapid modernization without clinical validation, governance, and operational discipline can create disruption that ultimately harms the very services technology is intended to improve.

The agency’s opportunity is to establish a stronger federal healthcare procurement model: one that rewards vendors for reliable EHR modernization, usable AI, interoperable data, resilient cybersecurity, and measurable improvement in veteran access and care coordination.

The contractors most likely to succeed will not be those that simply bring AI into every proposal. They will be the ones that can demonstrate how technology improves the daily experience of veterans, clinicians, and staff while remaining safe, secure, accountable, and practical to operate.