Building Capability, Not Just Buying Technology
The Public AI Brief · Issue No. 22
The federal workforce question finally got asked out loud this week: what happens to workers after AI takes over the mundane tasks? It’s the question agencies have been dodging, and the answer reveals whether AI becomes genuine capability transformation or just another efficiency-driven headcount reduction exercise.
This connects to a broader pattern I see across state and local governments. The most serious AI efforts prioritize organizational capability over technology acquisition. Georgia is training public employees statewide on AI literacy. Tennessee released a four-pillar action plan that treats workforce development and governance as equal priorities with pilots and infrastructure. El Paso created an AI apprenticeship program to build local talent rather than compete for expensive outside hires. The common thread is recognition that technology without organizational readiness consistently produces expensive failures.
Perhaps, they’re realizing what I’ve been saying for a while, that AI can be great for task completion, but it requires the deep human subject matter expertise as the key to unlocking the true potential. An organization of experts leveraging the power of AI will be light years ahead of short-sighted oragnizations who see AI simply as a means of reducing headcount.
This Week’s Key Developments:
Federal workforce transformation: Agencies confront what comes after AI automates routine work
Georgia launches statewide AI literacy initiative to train public employees across all agencies
Tennessee releases action plan with four strategic pillars including workforce development and governance
El Paso creates AI apprenticeship program building local government talent
AWS commits $50B to government AI infrastructure while data center resource concerns intensify
Federal
The conversation about federal workers and AI automation moved beyond displacement fears to capability transformation this week. Federal leaders acknowledge AI will handle routine tasks, but the harder question is whether agencies can successfully transition workers to higher-value activities. The challenge isn’t technological. It’s also not new. Even before ChatGPT came on the scene, senior leadership across the federal government has struggled to deploy the workforce to truly innovate, often settling for the tried and tested. The real question now is whether AI-promised efficiency gains become workforce development opportunities or just headcount reduction targets dressed up as modernization.
This connects directly to ongoing DOGE and U.S. Digital Service modernization work. DOGE continues as a temporary organization within USDS, with Amy Gleason as acting head, though long-term structure remains unclear. What matters more than organizational charts is the full slate of modernization projects underway, suggesting federal AI adoption continues regardless of political messaging around efficiency versus transformation.
Amazon Web Services announced $50 billion in AI and supercomputing infrastructure investment specifically for government customers, granting expanded access while scaling supporting infrastructure. The investment signals vendor confidence in sustained public sector AI demand, though it raises questions about whether agencies are building genuine capability or deepening cloud dependency. Thinking back to last month’s AWS outage, or last week’s Cloudflare failure, maybe we should think a little more about what the implications could be? Separate reporting on reducing data center water consumption highlights the environmental costs, suggesting infrastructure decisions must balance economic development incentives with resource constraints.
State
Building Organizational Capacity
Georgia is partnering with InnovateUS to train public employees statewide on working with AI. CIO Shawnzia Thomas emphasized that empowering people is central to digital transformation, not an afterthought. The literacy initiative recognizes what failed IT projects consistently demonstrate: technology adoption without workforce capability building produces expensive failures, not efficiency gains.
Tennessee released an AI action plan built on four strategic pillars: pilots, infrastructure, workforce development, and governance. The framework emphasizes organizational discipline over shiny tools, a refreshing contrast in a landscape dominated by vendor promises. The plan aims to modernize services and strengthen the economy, though success depends on whether the state can execute across all four pillars simultaneously or whether governance gets deprioritized when budget pressures hit.
New Mexico’s Health Care Authority launched a “digital front door” to streamline resident access to public assistance programs and call centers. The initiative focuses on customer experience and staff productivity simultaneously, avoiding the trap of optimizing one at the other’s expense. Like the Benefits Portal we have here in Maryland, I think success will depend on whether the technology actually reduces friction or creates digital barriers for residents with limited connectivity or digital literacy.
Workforce Challenges
State and local leaders emphasized “shared purpose” as essential for managing multi-generational workforces amid retirements and private sector competition. Agencies must appeal to civic duty while wrestling with AI’s role in workforce transformation. The challenge intensifies as AI potentially displaces some roles while creating demand for new capabilities that existing workers may or may not develop.
Leaders at the GOVIT Summit urged governments to “walk the talk” amid public distrust and volatility. Communication and keeping promises matter more than ambitious AI strategies that fail delivery. States navigating AI adoption while maintaining public trust need consistent execution, not just compelling vision statements.
Local
Workforce Development Models
El Paso partnered with SuperCity AI to create an AI apprenticeship program training applicants with AI skills for local government work. The initiative represents an alternative to traditional hiring, building capability from within the community rather than competing for experienced talent in expensive markets. Success depends on whether apprenticeships actually lead to government employment or just provide training that benefits private sector employers.
Long Beach wrapped workshops teaching residents digital skills including AI usage and how the city deploys it. The initiative combines resident education with transparency about city AI applications, recognizing that public trust requires understanding. Whether workshops reach beyond early adopters to residents with limited digital access remains an open question.
Governance Structures
New York City Council passed legislation creating an Office of Algorithmic Accountability to audit, monitor and regulate city agency AI tools. A separate initiative aims to educate the public on AI. This is NYC’s second attempt at AI regulation after previous efforts stalled, so the real test is whether this office gains actual enforcement authority or becomes another advisory body that agencies ignore when convenient.
Education
Warren County Community College in New Jersey launched an associate degree program combining five technology fields, including AI and robotics, into one curriculum. The program aims to prepare students for automation and manufacturing careers in a region where traditional pathways are disappearing. Community colleges often lead in pragmatic workforce development because they’re directly accountable to local employers and students who can’t afford credential programs that don’t lead to jobs.
Key Insights for Practitioners
Capacity building beats technology acquisition: Georgia’s statewide literacy initiative and Tennessee’s four-pillar framework both prioritize workforce development and governance alongside pilots and infrastructure. Technology without organizational capability consistently fails.
Action: Conduct an honest assessment of your organization’s AI literacy across all levels, not just technical staff. Invest in training before expanding deployments, even if that slows adoption.
Workforce transition planning is unavoidable: The federal question about what comes after AI automation applies to every government level. Organizations must decide now whether efficiency gains fund capability development or just become headcount reductions.
Action: Begin documenting where AI is replacing, augmenting, or transforming roles in your organization. Build transition plans that treat workforce development as essential infrastructure, not optional overhead.
Community-based workforce development works: El Paso’s apprenticeship model and Long Beach’s resident workshops show alternatives to competing for expensive outside talent. Building local capability creates sustainable capacity and strengthens community connections.
Action: Explore partnerships with community colleges, nonprofits, or workforce development organizations to create local talent pipelines. Focus on converting existing community members rather than recruiting from competitive markets.
What I’m watching: How agencies actually handle workforce transitions as AI deployments scale beyond pilots. If automation gains only fund more automation rather than worker development, expect unions and employee groups to push back hard on future AI initiatives.
How is your organization approaching AI workforce development? Are you investing in capability building before technology deployment, or running pilots first and figuring out the people side later? Share your experiences in the comments.