Introduction
As America’s AI Action Plan sharpens the federal government’s focus on artificial intelligence (AI), Federal Systems Integrators (FSIs) are stepping up as both internal innovators and trusted partners in public sector transformation.
With growing demand across cyber defense, digital services, and mission analytics, FSIs face a dual imperative: advance internal AI maturity while delivering secure, scalable solutions to government clients. Yet many organizations are still navigating the complexities of talent readiness, governance frameworks, and platform integration – slowing efforts to move AI initiatives from pilot to production.
In this new study, MeriTalk, in collaboration with Dell Technologies and NVIDIA, surveyed 75 IT leaders from FSI organizations to pinpoint where AI pilots are gaining traction and what differentiates projects that successfully scale.
For the purposes of this research:
- AI is the ability of computer systems to simulate human intelligence to perform tasks
- Generative AI (GenAI) is a subset of AI that leverages large datasets to create high-quality content in many forms, like text, images, video, and software code
- Agentic AI refers to emerging AI systems that can autonomously plan, make decisions, and take action to accomplish goals
Executive Summary
FSI leaders are gaining ground with AI, but pilot-heavy patterns and internal friction still block scale
- 96% of FSI IT leaders say AI is essential to delivering on their mission, and 89% already credit GenAI with measurable productivity gains
- Yet just 28% report scaling AI or embedding it across multiple workflows, and 95% say they’re more likely to start a new pilot than scale an existing one
- Most common factors in AI projects that have stalled or failed? Governance or compliance blockers (36%), integration challenges (33%), cybersecurity gaps (33%), and outdated infrastructure (31%)
Breakthrough FSIs are scaling AI with targeted investments in people, processes, and platforms
- When it comes to their people, FSI leaders say their most impactful steps in helping to scale AI have been giving staff time for AI exploration, improving change management practices, and embedding AI goals into leadership performance metrics
- On the process side, top enablers include establishing sandbox environments, creating formal governance frameworks, and standardizing AI lifecycle processes
- The most helpful platform-level changes include adopting scalable cloud or hybrid environments, modernizing data infrastructure, and deploying observability tools for continuous monitoring
Future success depends on talent, trust, and infrastructure
- FSIs say secure, scalable infrastructure and an AI-fluent workforce will be the primary drivers of AI success over the next three years
- Internally, they expect to see the greatest returns in workforce enablement (59%), project forecasting (49%), and cyber and supply chain resilience (47%)
- Externally, 83% report growing demand for agentic AI in public sector and see explainable AI, predictive analytics, and secure automation as critical to future contract wins
AI Momentum & Early Wins
FSIs are quickly moving beyond AI exploration, with the majority running AI pilots or working to scale adoption. Early efforts are paying off – 89% report productivity boosts and 96% call AI mission-critical – fueling gains in workforce output, code, proposal development, and knowledge flow.
More than half of FSIs are piloting or scaling AI:
- 43% of FSIs are exploring AI use cases
- 29% are running pilots or proofs-of-concept
- 28% report scaling or embedding AI across multiple workflows organization-wide
Momentum is growing:
- 89% say GenAI has already boosted productivity
- 96% call getting AI right “essential” to delivering on their mission
Where are FSIs seeing the most early success?
- 57% Boosting workforce productivity
- 51% Accelerating software/code development
- 45% Streamlining proposal workflows
- 43% Improving internal knowledge management
Why Good Pilots Stall
Despite progress, 96% of FSI leaders have watched at least one internal AI effort stall and estimate only 51% of client contracts move beyond pilots to scaled implementation. The most commonly cited roadblocks include governance logjams, tangled integrations, insufficient cyber planning, and outdated infrastructure.
- Internally, 96% of FSIs have had an AI project stall or fail
- Externally, just 51% of AI-related contracts move beyond pilots
Most common factors in AI projects that have stalled or failed?
- Governance or compliance blockers (36%)
- Difficulty integrating with existing systems (33%)
- Insufficient cybersecurity planning (33%)
- Outdated infrastructure or insufficient compute capacity (31%)
Additional Barriers to Scale
FSIs also report challenges with structural gaps and conflicting incentives. Fewer than half have formal governance (44%), performance KPIs (37%), or dedicated budgets (32%). Nearly all (95%) say their organization is more likely to launch a new AI pilot than scale an existing one.
Structural gaps:
- 44% have an approved AI governance framework
- 37% track formal KPIs to measure AI performance
- 32% have a dedicated AI budget separate from general IT
Culture and pressure:
- 95% are more likely to launch a new AI pilot than scale an existing one
- 80% hesitate to move even successful AI pilots into production due to perceived risk
- 80% feel pressure to show AI results – even if they’re not fully ready
What Works in FSI Organizations
Those effectively scaling AI credit efforts across three fundamentals: people, processes, and platforms. They pair robust change-management and exploration time with sandbox-driven processes and formal governance – then leverage cloud-ready, modern data stacks with real-time monitoring.
Which steps have been most important in helping to scale AI initiatives?
When it comes to our organization’s people:
- 51% Improving change management practices to drive AI adoption
- 51% Providing staff with time and space to explore AI tools
- 45% Embedding AI goals into leadership performance metrics
When it comes to our organization’s processes:
- 43% Establishing pilot programs or sandbox environments to test AI tools
- 40% Creating a formal AI governance framework
- 40% Standardizing AI lifecycle processes
When it comes to our organization’s platforms:
- 48% Adopting scalable cloud or hybrid environments for AI workloads
- 44% Modernizing data infrastructure to improve access and quality
- 41% Deploying tools for continuous AI monitoring and observability
The Changing Face of AI RFPs
When it comes to public sector AI, 83% of FSIs say they’ve noticed an increase in client interest in agentic AI, with 45% saying it’s now a common request in RFPs. As they consider future contracts, FSIs say ethical AI, predictive analytics, and AI-powered cyber capabilities will be critical.
- 83% have noticed an increase in public sector RFPs that include agentic AI, with 45% now saying it’s a “common requirement”
- 68% say sovereign AI – a nation’s effort to develop and produce AI using its own infrastructure and data – will be very important for driving innovation across the U.S. government and economy going forward
What new AI services or capabilities do you believe will be table stakes for winning future public sector contracts?
- Ethical, explainable, and transparent AI
- Predictive analytics and data-driven decision-making capabilities
- AI-powered cybersecurity and threat detection
- Personalized and citizen-centric services
- Automated processes and workflows
Looking Ahead: Success Drivers & Opportunities
Overall, FSIs say tomorrow’s AI wins hinge on tech and human talent: 44% single out scalable, secure infrastructure, while an AI-fluent workforce follows closely at 41%. Over the next five years, they expect the greatest internal AI returns in workforce effectiveness, on-time and under-budget project delivery, and cyber resilience.
What will be the primary factor that determines your AI success over the next three years?
- #1 Scalable, secure infrastructure that supports AI workloads and data needs (44%)
- #2 A skilled and AI-fluent workforce (41%)
- #3 A clear and actionable AI governance framework (15%)
Where do you see the greatest opportunity for AI to improve your organization’s internal operations over the next five years?
- 59% Enhancing workforce enablement and upskilling with AI copilots
- 49% Improving project delivery timelines and cost estimation accuracy
- 47% Strengthening internal cybersecurity and supply chain resilience
- 45% Scaling reusable AI models across business units
- 43% Increasing efficiency of internal data integration and analytics
Recommendations
Double down on AI fluency to elevate delivery.
FSIs rank an AI-fluent workforce nearly equal to infrastructure as the top success driver over the next three years. Equip staff not just to build and manage AI, but to guide clients on ethics, explainability, and mission fit. Offer role-based training for engineers, analysts, and client-facing teams – and embed AI proficiency goals into leadership performance reviews to hardwire accountability across the organization. Support workforce alignment with the AI Action Plan by contributing to public-private talent exchanges, delivering targeted AI skills training, and helping federal partners close capability gaps.
Move beyond the pilot trap with scalable architecture and KPIs.
With 95% of FSIs saying they’re more likely to launch a new pilot than scale a successful one, it’s time to reset incentives. Formalize KPIs tied to scale readiness, earmark funding for multi-phase deployments, and ensure every pilot includes a path to production. Invest in cloud-ready, modular infrastructure with observability tools and reusable pipelines to streamline AI rollout across teams and client environments. Help agencies meet their scaling goals by delivering solutions that align with governmentwide evaluation guidance and support open-source, open-weight models.
Codify governance that builds trust – internally and externally.
Governance and compliance blockers remain the No. 1 barrier to AI success for FSIs. Replace informal decision-making with transparent governance frameworks that define model performance thresholds, risk reviews, and lifecycle management. Clear governance boosts internal efficiency – and becomes a contract-winning differentiator in an environment where the administration is emphasizing explainability, risk mitigation, and secure-by-design AI.
Design to win the next generation of public sector work.
As public sector RFPs increasingly demand agentic AI, explainable models, and secure automation, FSIs must evolve their offerings. Build pre-approved solution sets aligned to these criteria, and develop frameworks that showcase traceability, predictive value, and zero-trust compliance. Position your team as not just a technology provider, but a strategic partner that understands the AI Action Plan’s priorities – supporting secure, auditable, and mission-ready AI that meets rising expectations for innovation and accountability.
Methodology
MeriTalk, in collaboration with Dell Technologies and NVIDIA, surveyed 75 federal IT decision-makers from FSI organizations in June 2025. The research has a margin of error of ±11.29% at the 95% confidence level.
| Job Title | |
|---|---|
| CIO, CTO, CAIO, or CISO | 45% |
| Deputy or assistant CIO, CTO, CAIO, or CISO | 15% |
| Director of IT/technology | 34% |
| IT program manager/officer | 1% |
| Data, AI, infrastructure, or network manager | 5% |
100% of respondents are familiar with their organization’s current or planned use of AI