Jane Edwards 
- Cloudera executive Dario Perez has outlined data sovereignty principles for AI adoption
- AI governance should extend across the full lifecycle
- The 2026 FedCiv Summit will examine AI, cloud, cybersecurity and more
Dario Perez, GVP, federal civilian and state, local and education at Cloudera, said organizations in the public sector and other regulated industries should approach artificial intelligence adoption with data sovereignty in mind by designing AI strategies that respect existing data constraints instead of adapting sensitive data to fit AI tools.
As agencies continue to adopt AI while managing security and compliance requirements, government and industry leaders will discuss priorities shaping civilian agencies at the Potomac Officers Club’s 2026 FedCiv Summit on Oct. 29. The event will feature discussions on AI adoption across government, data and cloud infrastructure, cybersecurity and cross-agency programs. Register now!
In a company blog post published June 24, Perez wrote that organizations should account for where data resides, how it is governed and which regulatory requirements apply, adding that this approach is particularly important in regulated industries. He also said data sovereignty depends on how data, governance and AI systems operate together across environments.
What Is Data Sovereignty?
Perez described data sovereignty as an organization’s ability to maintain control over its data, including which laws and regulations apply, who can access the data and how it can be shared, used and protected. He noted that related concepts such as data residency, data localization and data governance support data sovereignty but do not independently establish it, describing sovereignty as “sustained, enforceable control under a defined legal and regulatory framework.”
What Principles Support Data Sovereignty in AI?
Perez said organizations seeking to maintain data sovereignty while deploying AI should focus on several core principles:
- Bring AI to the data — AI workloads should operate close to where sensitive data resides, whether on-premises, in secure cloud environments or across distributed systems, reducing unnecessary data movement and helping organizations maintain greater control.
- Apply governance across the AI lifecycle — Governance should extend beyond datasets to cover model training, deployment, reuse and retirement while applying access controls, lineage and usage policies to AI models and derived artifacts.
- Maintain visibility, lineage and accountability — Organizations should establish transparency into data origins, model training and AI outputs to support audits, regulatory compliance and public accountability over time.
Why Does Perez View Data Sovereignty as an Ongoing Discipline?
Perez said organizations should no longer view data sovereignty as a one-time compliance activity but as an ongoing discipline that influences how AI systems are designed, deployed and governed. He noted that Cloudera supports public sector and other regulated organizations in developing AI capabilities while addressing data sovereignty requirements.
The discussion of secure, governed AI deployment extends to a recent Executive Spotlight interview with Robert Gilman, senior account executive for the Department of War at Cloudera. During the interview, Gilman discussed how agencies are navigating the evolving AI landscape, including the role of air-gapped environments, private AI, modular architectures, governance frameworks and secure deployment models in supporting next-generation government AI initiatives.
