Beyond Empowerment: The Case for a Veterans AI Access and Equity Act

I. Introduction

In January 2024, I wrote an article titled “Empowering NYC Veterans: Using Artificial Intelligence for a Brighter Future,” which explored how Artificial Intelligence (AI) tools could be incorporated into a legal strategy that addresses veteran poverty in New York City. Approximately one year later, following Executive Order 14179, the U.S. Department of Veterans Affairs (VA) launched an AI initiative aimed to “improve outcomes and experiences for our Veterans by developing trustworthy AI capabilities in support of the agency’s mission.”[1] Through this initiative, the VA promises to improve healthcare outcomes for veterans and to streamline access to veterans’ benefits by using AI to enhance efficiency and prevent fraud. While these actions are important, there is a mismatch between the VA’s goals and the daily needs of U.S. veterans.

The AI tools the VA has begun to employ focus inward, making VA processes faster and helping VA staff navigate paperwork more efficiently.[2] However, this neglects the basic needs that push 1.2 million veterans into poverty such as housing, employment, and mental health challenges.[3] Therefore, over a year later, a law that mandates the availability of AI-powered career readiness tools and AI-powered mentorship networks is still necessary. This law could be structured like a federal Veterans AI Access and Equity Act, requiring agencies to use AI to meet the core obligations of the GI Bill and the Uniformed Services Employment and Reemployment Rights Act of 1994 (USERRA).

II. Existing Structural Barriers

In New York City alone, more than 20,000 veterans live below the poverty line.[4] The 1944 GI Bill was designed to provide access to higher education, job training, and financial benefits.[5] USERRA was enacted to protect reemployment rights for returning service members.[6] In 1979, Callahan v. Carey established a constitutional right to shelter under the New York State Constitution.[7] However, these laws have not prevented systemic failure.

On the city level, a significant number of violations of the Callahan ruling have occurred, including insufficient shelter capacity, denial of stable shelter placements, and no action by the city regarding flooding of sleeping areas.[8] Furthermore, New York City has attempted to limit the Callahan decision by denying shelter to homeless people because of their non-compliance with social services, and by attempting to change eligibility rules to deny shelter to thousands of people suffering from mental and physical illnesses.[9] On the national level, veterans continue to struggle to obtain their legal benefits not because the laws do not exist, but because the systems built to enforce them are fragmented and inconsistent.

III. The U.S. Department of Veterans Affairs AI Initiative

Executive Order 14179 intended to remove “barriers to American AI innovation, clearing a path for the United States to act decisively to retain global leadership in artificial intelligence.”[10] When announcing its AI initiative, the VA stated that, consistent with the Executive Order, it is “investing in AI solutions that support the delivery of world-class benefits and services to veterans.”[11] Some of those benefits include “advancing health equity…, making health care more affordable, improving outcomes through more coordinated care, improving patient experience, and reducing clinician burnout,” as well as applying AI tools to “areas such as suicide prevention, medical diagnostics, fraud detection, and internal efficiency measures.”[12] On its face, this initiative is an overdue attempt to bring advanced technology into a fragmented system.

However, the steps taken pursuant to this initiative have focused almost entirely on internal operations. For example, VA’s AI inventory includes the Evolv Weapons Detection System, used internally to create “visual representations of data sets for reports and presentations.”[13] The VA has also begun using the “Avigilon Camera/Search Function,” which allows VA employees to search for information using AI.[14] A third item in the AI inventory is eCaremanager, which “integrates clinical, physiological, and demographic data from all patients” and presents the data graphically to VA employees.[15] While these are important goals, the initiative reveals a clear pattern: the technology is being used to refine internal operations, not to expand the rights of veterans.

IV. Career Readiness and Employment Tools

To meaningfully affect U.S. veterans’ daily lives, the use of AI in veterans services needs to shift from administrative convenience to real-world applications. Many veterans return from their service with valuable skills but struggle to find employment due to inadequate job networks.[16] Therefore, agencies like the VA could use AI to identify growing job markets across the country and provide customized skills training and certificate programs. For instance, veterans can effectively contribute to the technology industry with the attention to detail, accuracy, and record-keeping skills taught in the military. AI systems can be used to create lesson plans, teach veterans, and track their progress, allowing the government to assist veterans without incurring excessive costs. This solution is plausible because dozens of AI-powered career readiness tools already exist, including an AI-powered career transition support tool designed specifically for veterans: Redeployable is a tool intended to make it easier for veterans to “navigate the job market, pinpoint where they fit, and secure a role that actually aligns with their skills and ambitions.”[17] To do so, Redeployable translates skills and goals into personalized career recommendations, provides structured road maps to transition into that career, and even identifies available job openings.[18] Therefore, as part of a larger veterans rights law, agencies like the VA should be mandated to provide services like this to ensure that they are free and available to every veteran.

V. AI-Powered Mentorship Networks

In addition to career tools, peer-led groups and mentorship networks should be mandated to reach the entire veteran population. Mentorship networks can help connect veterans with employers, professionals, and other veterans who can provide guidance in their transition to civilian life. Peer-led groups can also help reduce addiction among veterans, ultimately reducing homelessness and poverty. However, peer-led groups are often unsuccessful because they are largely self-organized or assigned randomly. Thus, individuals can easily mimic the negative behaviors of peers around them, causing their behavior to worsen.[19] To solve this problem, agencies like the VA can use AI to group veterans into efficient support groups. The AI software could account for positive and negative relationships, mathematically maximize the number of existing friendships, and minimize the number of substance abusers by putting them in positive situations that encourage recovery.

VII. Conclusion

The VA’s current AI initiative may make the agency itself more efficient, but it has not done enough for veterans in their daily lives. Poverty, unemployment, and untreated mental health issues are still widespread among veterans in New York and across the country. These issues are not new, and many laws meant to address them already exist; however, no law to date mandates the availability of AI-powered tools for career readiness and mentorship. The VA is beginning to propose cutting-edge solutions. However, AI will be more effective if it is directed outward, toward real-world outcomes, not just to internal operations. Therefore, a Veterans AI Access and Equity Act should be enacted to require agencies to use AI to deliver on the promises already made to veterans: job access, housing, health care, and effective support systems.

[1] “VA Artificial Intelligence.” U.S. Department of Veterans Affairs, department.va.gov/ai/.

[2] “VA AI Use Case Inventory.” U.S. Department of Veterans Affairs, department.va.gov/ai/ai-use-case-inventory/.

[3] “S2101 Veteran Status.” United States Census Bureau, 2023, data.census.gov/table?q=S2101.

[4]  “New York City Veterans Demographics.” NYC Veterans. www.nyc.gov/assets/veterans/downloads/pdf/NYCveterandemographics.pdf.

[5]  “G.I. Bill.” History.com, www.history.com/topics/world-war-ii/gi-bill#section_2.

[6] Uniformed Services Employment and Reemployment Rights Act of 1994, 38 U.S.C. §§ 4301-4335 (1994).

[7] Callahan v. Carey, 53 A.D.3d 404, 861 N.Y.S.2d 624, NY Slip Op 05968 (App. Div. 2008).

[8] “The Callahan Legacy: Callahan v. Carey and the Legal Right to Shelter.” Coalition For The Homeless. www.coalitionforthehomeless.org/our-programs/advocacy/legal-victories/the-callahan-legacy-callahan-v-carey-and-the-legal-right-to-shelter/.

[9] Ibid.

[10] Exec. Order No. 14,179, 89 Fed. Reg. 45,885 (June 24, 2024).

[11] “VA Artificial Intelligence.” U.S. Department of Veterans Affairs, department.va.gov/ai/.

[12] “VHA Aligns with Leading Health Care Organizations to Ensure Trustworthy Use of AI.” U.S. Department of Veterans Affairs, 14 Feb. 2024, department.va.gov/ai/vha-aligns-with-leading-health-care-organizations-to-ensure-trustworthy-use-of-ai/.

[13] “VA AI Use Case Inventory.” U.S. Department of Veterans Affairs, department.va.gov/ai/ai-use-case-inventory/.

[14] Ibid.

[15] Ibid.

[16]  U.S. Department of Health and Human Services, Office of Human Services Policy, “Toward Understanding  Homelessness: 2007 National Symposium on Homelessness Research: Housing Models,” aspe.hhs.gov/reports/toward-understanding-homelessness-2007-national-symposium-homelessness-research-housing-models-0.

[17] “Cracking the Code: How AI Helps Veterans Find the Right Civilian Career.” CivvyJobs, 23 Mar. 2025, civvyjobs.com/resources/knowledge-hub/2025/03/cracking-the-code-how-ai-helps-veterans-find-the-right-civilian-career/.

[18] Ibid.

[19] TEDxTalks. “How Social Workers Use AI to Help Unhoused Teens | Anamika Barman-Adhikari | Tedxmilehigh.” YouTube, 13 Oct. 2022, www.youtube.com/watch?v=1E7BoJQf050.