Empowering People with Sovereign Agents: A National Human + Agent Model for the Agentic Economy
Most conversations about agentic AI start with autonomy: how to remove people from the loop and automate work. That path may improve short-term efficiency, but it risks displacing human value as models and agent stacks become widely available.
This talk proposes a different national trajectory: empower people with sovereign agents that act as lifelong digital extensions of their identity, skills, and judgment. Instead of competing with autonomous systems, citizens, public servants, and enterprises would operate through a Human + Agent model where humans remain accountable, agents expand capability, and sovereignty is designed into the architecture.
The session introduces OSARA (Open Sovereign Agent Reference Architecture) as an open implementation path for countries and institutions: identity, control, interoperability, evidence, and governance mechanisms that help humans stay in the economy, excel within it, and outperform systems optimized only for full autonomy.
Countries are being told the future of AI is fully autonomous. In practice, that may be the wrong north star.
When foundation models converge and agent frameworks become common infrastructure, autonomy alone stops being a durable advantage. What remains scarce is judgment, trust, accountability, and the ability of people and institutions to use AI on their own terms.
This talk argues that governments and national digital ecosystems should do something more ambitious than deploying autonomous agents:
- Empower people with sovereign agents they can trust
- Treat those agents as lifelong digital extensions of human capability
- Build the surrounding mechanisms for identity, authorization, evidence, and interoperability
- Help humans remain economically relevant, productive, and competitive as agentic AI scales
The goal is not to slow AI down. It is to ensure that national AI strategies amplify people rather than replace them.
The session then presents OSARA, the Open Sovereign Agent Reference Architecture, as a practical open framework for implementing this model at country level and downstream across ministries, enterprises, education systems, and public services. OSARA is designed around Bring Your Own Agent (BYOA), Human + Agent pairing, and verifiable control so organizations can adopt powerful agents without surrendering ownership, accountability, or sovereignty.
Audience: policymakers, CIOs, CISOs, digital transformation leaders, open-source practitioners, and architects designing national or institutional AI strategies.
Mostafa Elkordy brings over two decades of hands-on experience in software architecture and large-scale system development, with a focus on secure, interoperable, and resilient digital systems in complex international environments. His work spans enterprise integration, identity and access management, and security assurance, including adversarial validation and penetration testing of large-scale systems. He currently serves as Senior Cybersecurity Solutions Officer at UNICC and as Chair of Open Source United, a Community of Practice within the United Nations ecosystem.
He has led initiatives in secure platform engineering and digital trust, with a strong emphasis on practical approaches to cybersecurity, system resilience, and cross-organizational interoperability.
He is also the founding author of OSARA (Open Sovereign Agent Reference Architecture) and its companion AI Bill of Rights, an emerging open framework intended to provide holistic guidance on how individuals, enterprises, and governments can interact through AI agents in a sovereign, secure, and interoperable manner. The work also explores broader implications for the future of work, as well as the regulatory and governance considerations required to support agent-driven digital ecosystems.