This post continues a series by Sarjoo Shah, Client Executive Director at RELI Group. In Part 1, How AI Is Transforming Program Integrity and Fraud Prevention in Government, Sarjoo examined the evolving landscape of Program Integrity – how increasing data complexity, sophisticated fraud schemes, and shifting regulatory expectations are reshaping oversight across all levels of government, and how innovations like AI can empower organizations fighting back against fraud, waste and abuse (FWA). In this second installment, we turn our focus to the state and local marketplace, where many of the federal lessons learned can serve as a powerful guide.
By adapting proven strategies in prevention, data infrastructure and governance, state and local agencies can modernize their approach to fraud, waste, and abuse (FWA) – achieving not only greater efficiency and accountability, but also building the foundation for sustainable, data-driven integrity programs that protect public trust and maximize taxpayer value.
Applying Federal Lessons to the State & Local Marketplace
The federal experience in program integrity offers a valuable playbook. For leaders in strategy and business development for state and local governments, here are three key areas to apply or adapt.
Lesson 1: Shift from “Pay-and-Chase” to “Prevent-and-Prioritize”
Federal programs have moved toward prioritizing leads, deploying predictive analytics to score risk, automating initial reviews and triaging investigative efforts.
For state/local agencies, adopting a similar mindset – early detection, risk scoring and workflow automation – can yield major efficiency gains. Instead of waiting for an audit queue to build, build systems that proactively surface high-risk transactions and allow for resource prioritization.
Lesson 2: Invest in Data Foundations and Ecosystem Infrastructure
Many federal efforts stall not because of a lack of algorithm ideas, but because of data fragmentation, inconsistent formats or weak linking across datasets. The federal mindset is pushing toward data-sharing frameworks, interoperability standards and clearing houses.
For state and local governments, this means cleaning and linking data like vendor lists, payments or service records; building modern cloud or hybrid data pipelines; and investing in tools that allow analytics rather than just static reporting. It means thinking not just about “fraud detection” but about “data-driven oversight.”
Lesson 3: Build Governance Frameworks that Align With Operational Change
Technology alone doesn’t deliver results. Federal agencies recognize that governance, accountability, workforce training, processes and organizational culture matter.
In state and local government, this means designing a holistic program integrity strategy that includes leadership accountability with a designated senior official responsible for FWA detection, cross-agency collaboration, continuous training, defined workflows for analytics-to-investigation hand-off, and regular measurement of outcomes (e.g., reduction in improper payments, speed of investigations, recovery rates, etc.).
Moreover, working at the state and local level requires a tailored approach with fewer dedicated investigators, more reliance on outsourced analytics or shared services, and tighter alignment with program owners (not just auditors).
A Forward-Looking View: What’s Next for State & Local Program Integrity
Looking ahead, here are some of the strategic imperatives and emerging possibilities for state and local government program integrity.
AI Embedded in Operational Workflows
Rather than stand-alone analytics tools, AI will increasingly be embedded into the lifecycle of payment, awarding, procurement and monitoring. This will likely impact systems like vendor onboarding processes that automatically score risk, real-time payment authorization systems that integrate analytics, or grant-monitoring dashboards that flag programmatic anomalies. This means oversight becomes continuous and nearly real-time.
Hybrid Human/AI Decision Models
Oversight teams will shift from being purely reactive investigators to “human-in-the-loop” models, where AI triages, scores and proposes leads, and analyst teams validate and act. This “augmented workforce” model increases throughput and prioritization of complex cases. Lessons from healthcare FWA show this shift is already happening.
Expanding to New Domains Outside Traditional High-Volume Claims
While a lot of focus has been on federal healthcare programs, state and local governments can apply these techniques in other domains, such as procurement fraud, inter-agency grants, disaster relief funds, social services eligibility, payroll/time‐keeping abuse, or infrastructure grant programs. The innovations will migrate horizontally.
Generative AI, Synthetic Data and Privacy-Preserving Analytics
As agencies gather more sensitive data and partner with vendors or third parties, privacy and security becomes even more critical. Emerging technologies like synthetic data for training AI models, federated learning involving sharing insights without sharing raw data, and advanced anonymization will open new avenues for analytics while safeguarding privacy. At the same time, generative adversarial techniques may be used by fraudsters, so agencies must stay ahead of that risk.
Cross-Jurisdictional Collaboration and Shared Services
Fraud rings increasingly cross state and local boundaries. Building shared analytics platforms, data clearinghouses, and cooperative task forces will become necessary. States can pool resources and share vendor risk databases, and local governments can leverage state or regional centers of excellence for analytics. The federal push toward clearinghouses serves as a model for this type of cross-jurisdictional collaboration.
Ethical, Transparent and Auditable AI governance
As use of AI grows, so do demands for transparency, fairness and accountability. State and local agencies will need to establish policies and frameworks that ensure decisions made by algorithms are explainable, models are monitored for bias or degradation, and audit trails exist. Recent research outlines the so-called “AI Fraud Diamond,” which highlights the need to think about algorithmic misuse or systemic bias as part of fraud risk.
Measurement, Outcomes and Public Trust
Ultimately, program integrity is as much about public trust as it is about cost savings. Agencies that demonstrate measurable results – reduced improper payments, shorter investigation times, better coordination and fewer abusive practices – will justify investment in analytics programs. Transparent reporting of these metrics can reinforce legitimacy and help secure budget support for continued innovation.
Recommendations for State & Local Leadership
Given these trends and innovations, here are actionable recommendations for leaders at state and local agencies:
- Start with a strategic blueprint for FWA risk, analytics and AI. Map your key program domains, identify high-risk flows (payments, grants, procurement, eligibility), assess current analytics maturity and data readiness, and define where AI/ML can deliver the highest value.
- Invest in data foundations and analytic enablement. Prioritize data-integration: vendor/contractor registries, payment data, claims/eligibility data, previous audit outcomes, etc. Implement data governance frameworks and build the analytics environment – cloud or hybrid – with scalability and flexibility.
- Pilot smart analytics in one high-value domain. Rather than “boil the ocean,” pick one program area with manageable scope and high risk (for example, a major grant program or large payment flow). Use AI/ML and anomaly detection to triage leads, automate workflows and measure results. Use the insights to build a business case for wider rollout.
- Build cross-agency collaboration and shared services. Create mechanisms to share data and insights across program divisions and agencies (state + local). Consider vendor risk databases, cross-jurisdiction taskforces or shared analytics platforms. Use inter-agency MOUs, governance forums and joint investigative workflows.
- Embed governance, transparency and workforce readiness. Ensure that oversight of analytics models includes periodic review, auditing, bias monitoring, algorithm explainability and documented decision-making. Develop training for staff in analytics and AI literacy. Ensure senior leaders are accountable for FWA programs and analytics adoption.
- Embed analytics into operations and workflow. Change the mindset: analytics is not just a reporting tool – it should be built into the program lifecycle. For example, vendor onboarding, payment authorization, contract amendments and grant monitoring should all include risk scoring and automated alerts.
- Measure outcomes and communicate value. Define key performance indicators (KPIs), such as reductions in improper payments, time to detection/lead resolution, number of investigative leads prioritized, recoveries achieved, cost per investigation, etc. Publicize results internally to drive investment, and externally to build trust.
- Plan for scaling, continuous learning and future threats. Fraud schemes evolve, so your analytic models must evolve. Build feedback loops; investigative outcomes feed model tuning. Monitor model performance degradation. Anticipate emerging threats, such as generative AI-enabled fraud, synthetic identities or cross-border schemes, and build adaptive capabilities.
Conclusion
For state and local government leaders, program integrity is now a strategic imperative, not simply a compliance or audit afterthought. The volume and complexity of funds flowing through programs, combined with increasingly sophisticated fraud, waste and abuse schemes, mean that traditional oversight approaches are no longer sufficient.
Artificial intelligence, advanced analytics and modern data architectures offer a pathway to shift from reactive to proactive, from undifferentiated investigation to targeted risk prioritization, or from manual review to intelligent automation. But technology alone doesn’t ensure success. Data readiness, governance, cross-agency collaboration, operational workflows and workforce change are equally critical.
As leaders build strategy and business development for state and local government programs, the time to act is now. Start small and strategic, show early wins, build the foundation for scale, and anchor analytics into every payment, grant, procurement and oversight process. Taxpayers expect nothing less – and the rising cost of inaction is real.
