The transformation of AI in the public sector: five key changes for concrete results
The public sector is experiencing an unprecedented operational revolution. If in the past artificial intelligence (AI) was mainly a tool to improve individual productivity, today the conversation has changed radically. Public bodies are moving beyond the isolated use of AI for specific tasks, such as document synthesis or workflow planning, to complex solutions that orchestrate complete organizational processes.
Agentic AI is becoming fundamental to solving real problems, improving citizen services, and detecting fraud and risks. This transition requires a deep reorganization not only of algorithms but also of data architecture, governance frameworks, and the relationship between human experience and automation.
Investing in AI for measurable results
Public agencies are allocating a significant portion of their operational budgets to ensure these new capabilities produce tangible results. According to Massimiliano Claps, research director of IDC, many U.S. public agencies expect to allocate 11% or more of their IT budget to AI by 2026. This investment is driven by the need to directly link technologies to service innovation and operational continuity.
About 80% of the agencies surveyed expect measurable results within 12 months and a return double the investment within two years. These ambitious goals require a unified platform that offers immediate visibility across the entire agency's operations, leaving no room for isolated data silos.
Rethinking architecture for scalable impact
Testing AI on a few PDF files is simple, but scaling these solutions across the entire organization requires an open and flexible architecture. Dave Erickson, a distinguished architect in the public sector at Elastic, emphasizes the importance of avoiding new data silos tied to a single cloud or vendor. OpenTelemetry emerges as a key tool to ensure the level of agnosticism necessary for data integration.
The architecture must be designed to enable fast and accurate searches across petabytes of information. James Garside, senior customer enterprise specialist at Elastic, observes that data cannot simply be stored in static buckets and expect AI to automatically extract useful information. For example, the UK took time to properly implement these solutions, demonstrating that discipline and planning are essential.
Redefining control: from human-in-the-loop to human-on-top
In a recent survey conducted by IDC, none of the 152 U.S. federal IT and mission leaders interviewed said they wanted to completely abandon human oversight. This data is significant: the role of humans in automation is evolving from the "human-in-the-loop" model, where the analyst merely approves decisions, to a "human-on-top" approach, where AI acts as a dedicated assistant.
In this new paradigm, AI manages data at scale, while humans define the strategy and make the final decisions. Erickson emphasizes that AI should help automate already established processes, keeping humans in control and ensuring that institutional knowledge remains the cornerstone of every decision.
Strategic autonomy and the imperative of sovereign AI
As the importance of data as fuel for agentic workflows increases, control over where data resides and who can access it becomes crucial. Sovereign AI is emerging as a global priority for organizations handling sensitive or classified information.
According to IDC, 46% of the federal entities surveyed already use some form of sovereign AI, and another 38% plan to adopt it within the next 12 months. Sovereign AI is not about isolation but control over where data resides, how it is exchanged, and who can access it with what rights. To maintain independence in their technological architecture, IT leaders must ensure control over their technological stack, starting from the fundamental data level.
Prioritizing next steps for AI integration
The transition from individual use of AI to mission-critical impact is already underway. To keep up, organizations must move beyond isolated experiments and focus on the architectural foundations that support AI at scale.
Starting an audit of the current data landscape to identify silos that hinder real-time access is the first step. Establishing governance frameworks that prioritize a "human-on-top" operational model ensures that teams maintain control over critical decisions. Finally, investing in a flexible platform that enables visibility and maintains strategic autonomy over the most sensitive data is essential.
Connecting distributed data to the experts who use it daily is the key to unlocking the true value of AI. To discover how to build this foundation for mission-critical success, you can participate in the dedicated webinar.
The impact of sovereign AI and the future challenges of the public sector
The adoption of AI in the public sector is not limited to improving operational efficiency but represents a strategic transformation that touches fundamental aspects such as data sovereignty and information governance. In an increasingly interconnected global context, the ability to ensure control over sensitive data becomes crucial for national security and technological independence.
The challenge of cybersecurity and defensive AI
With the increase in cyber threats, sovereign AI plays a key role in protecting critical infrastructures. Organizations must implement defensive AI solutions capable of detecting and neutralizing attacks in real time. This requires a proactive approach where AI not only analyzes historical data but predicts and prevents vulnerabilities before they can be exploited.
An emblematic example is the case of the UK, which has invested in defensive AI systems to protect government networks. However, as James Garside observes, "it is necessary to have a flexible regulatory framework that can evolve with technology while ensuring the protection of citizens."
The importance of training and education
One of the most significant obstacles to the adoption of AI in the public sector is the lack of specialized skills. Many public employees are not prepared to work with advanced technologies such as agentic AI. To bridge this gap, organizations must invest in continuous training programs and collaborate with universities and research centers.
An effective approach is the one adopted by the United States, where the government has launched initiatives to train public employees in the use of AI. These programs not only improve technical skills but also promote a culture of innovation within government agencies. As Dave Erickson emphasizes, "training is essential to ensure that employees can fully leverage the potential of AI while maintaining control over critical decisions."
The future of AI in the public sector
The future of AI in the public sector is promising but requires a strategic and collaborative approach. Organizations must move beyond isolated experiments and focus on the architectural foundations that support AI at scale. This includes adopting governance frameworks that prioritize a "human-on-top" operational model and investing in flexible platforms that ensure visibility and strategic autonomy over the most sensitive data.
Final thoughts
The adoption of AI in the public sector represents an unprecedented transformation, touching fundamental aspects such as data sovereignty, cybersecurity, and integration with emerging technologies. To keep up with these developments, organizations must invest in training, collaborate with experts, and adopt a strategic approach that ensures control over critical decisions.
In an increasingly interconnected world, sovereign AI emerges as a global priority, allowing organizations to maintain independence in their technological architecture. With the right strategy, the public sector can fully leverage the potential of AI, improving services for citizens and ensuring the security of critical infrastructures.
Editorial Note and Disclaimer
The guides and content published on GoYou are the result of independent research and analysis activities, for informational, educational, and in-depth purposes.
GoYou does not constitute a journalistic publication or an editorial product pursuant to Law No. 62/2001 and does not provide real-time information.
The GoYou project does not provide professional, technical, legal, or financial advice and disclaims any liability for the improper use of the information published.
In the Crypto sector, every investment involves risks: readers are invited to always inform themselves autonomously before making any decision.