Google targets ‘agentic workforce’ with Gemini for government push

As artificial intelligence moves from experimentation into production, government agencies are becoming an unexpected proving ground for large-scale workforce transformation. Faced with mounting operational demands, aging application environments and strict compliance requirements, public sector organizations are increasingly turning to AI not as a future capability, but as an immediate force multiplier — accelerating the shift toward an agentic workforce.

That shift is central to Google Cloud’s latest push around Gemini for Government, which positions AI as both a productivity layer and a modernization engine across agencies.

In the latest episode of theCUBE Research’s AppDevANGLE podcast, Principal Analyst Paul Nashawaty spoke with Chris Hein, field CTO at Google Public Sector, about how agencies are adopting AI, navigating compliance complexity and beginning to define what an “agentic workforce” looks like in practice.

“I think what we’re going to see … is really starting to see the agentic workforce take place,” Hein said. “Where it’s a valuable part of the team.”

The agentic workforce moves from concept to reality

Much of the enterprise AI conversation over the past two years has centered on pilots, proofs of concept and chatbot-style interfaces. In the public sector, the focus is shifting toward something more immediate: helping workers do more with less — a foundational step in building an agentic workforce.

Hein described a phased approach that begins with augmenting existing workflows rather than replacing them outright. That includes using AI to streamline administrative tasks, improve access to information and increase day-to-day productivity for government employees.

This aligns with broader market data. According to theCUBE Research, 46.5% of organizations report pressure to deliver applications significantly faster than they did three years ago, while headcount growth has not kept pace. AI is increasingly viewed as the only viable way to close that gap and enable an agentic workforce.

For agencies, the key is creating environments where employees can safely experiment with AI tools while remaining within compliance boundaries.

Compliance becomes the gating factor

Unlike commercial enterprises, where innovation can often move ahead of governance, public sector AI adoption is constrained, and defined, by compliance requirements. Security accreditation, data residency, privacy controls and regulatory alignment are not secondary considerations; they are the starting point.

Google’s approach, as Hein described it, is to embed those controls directly into its cloud platform rather than isolating AI capabilities into separate, restricted environments. The result is a model where agencies can access frontier AI capabilities while maintaining required standards such as FedRAMP and Department of Defense compliance — a necessary foundation for scaling an agentic workforce.

This dynamic is increasingly relevant beyond government. As regulatory frameworks such as the EU Cyber Resilience Act come into force, developers across industries are facing similar pressures to ensure that applications (and now AI systems) are compliant by design.

Modernization shifts from project to process

Legacy modernization remains one of the most persistent challenges in government IT, with many agencies still operating mission-critical systems built on decades-old architectures. What’s changing is how modernization is approached.

Hein pointed to growing use of AI-assisted development to help refactor and modernize systems, including efforts to update COBOL-based applications that have long resisted transformation.

Rather than treating modernization as a one-time migration event, agencies are beginning to view it as a continuous process, supported by AI-driven tooling that improves code, documentation and workflows incrementally — reinforcing the operational backbone of an agentic workforce.

This reflects a broader shift across application development: modernization is no longer a destination. It is an ongoing discipline.

Open models and optionality gain importance

Another key theme is flexibility. Government agencies, like large enterprises, are wary of long-term dependence on any single AI provider or model. Hein emphasized the importance of offering access to a range of models, including open-weight and frontier options, through platforms such as Vertex AI.

“I think it’s really important for government to have optionality,” he said.

This multi-model approach allows organizations to optimize for performance, cost, security and use case requirements, while maintaining the ability to adapt as the AI landscape evolves — a critical capability for sustaining an agentic workforce.

For developers, this suggests a future where applications are designed to interact with multiple models rather than being tightly coupled to one.

Edge deployment expands the AI footprint

AI workloads are also becoming more distributed.

Hein highlighted the need to tune and deploy models at the edge, enabling agencies to run AI in environments where latency, connectivity or operational constraints make centralized inference impractical.

This introduces new complexity for developers and platform teams, who must now manage model lifecycle, governance and deployment across both centralized and edge environments.

It also reinforces the idea that AI is not just a cloud service; it is an architectural layer that spans the entire application environment — and supports the expansion of an agentic workforce.

The bottom line

Government adoption of AI is often assumed to lag behind the private sector. In this case, the opposite may be true.

Driven by workforce constraints and mission-critical demands, agencies are moving quickly to operationalize AI in ways that directly impact productivity, application development and service delivery.

The concept of an “agentic workforce,” where AI systems act as task-oriented collaborators, is emerging as the next phase of this evolution.

For developers, the implications are clear: AI is no longer just a feature. It is becoming part of the workforce, the platform and the application architecture itself — defining the future of the agentic workforce.

Here’s the full conversation with theCUBE Research’s Paul Nashawaty and Chris Hein, part of the AppDevANGLE podcast series:

Image: SiliconANGLE