Sovereign cloud security and AI data residency

Artificial intelligence has moved firmly to the centre of national economic and technological agendas.

In regions like the United Arab Emirates, government entities and highly regulated industries are rapidly adopting AI across critical services to drive innovation, operational efficiency, and public sector transformation. However, as these advanced technologies embed themselves deeper into core operations, the need to protect sensitive information has never been greater. For many nations, this means establishing sovereign cloud environments where data remains strictly within national borders, subject to local laws and governance. Maintaining this digital sovereignty is critical, yet it presents a complex, ongoing challenge for organisations that must continually balance the transformative benefits of cloud-based AI with stringent, evolving data residency requirements.

The escalating cyber security landscape

As digital transformation accelerates, the attack surface widens proportionally. Cyber threats are becoming more sophisticated, frequently targeting critical national infrastructure. Recent figures from the UAE Cyber Security Council illustrate this growing risk, revealing that nearly 750,000 cyberattack attempts targeting vital sectors, particularly energy and education—have been thwarted in recent months. To defend against these high-volume, increasingly complex attacks, security teams require the immense processing power and analytical speed of advanced AI.

However, utilising traditional cloud-based AI security tools often involves sending telemetry and security logs outside the sovereign boundary to external servers for analysis. This creates a significant compliance hurdle. According to a 2026 study by the IBM Institute for Business Value, 74% of surveyed UAE executives find meeting data residency and sovereignty requirements challenging.

Balancing advanced threat detection with data control

To solve this dilemma, the industry is shifting towards a model where the AI is brought directly to the data, rather than exporting the data to an external AI model. This approach allows highly regulated sectors to leverage machine-speed threat detection without compromising their sovereign boundaries. By running open models entirely inside customer-controlled infrastructure, organisations can process vital security signals, identify intricate attack paths, and automate defensive responses natively within their own secure environments, removing the need for external data transit.

Recognising this crucial market need, TrendAI™ has expanded its collaboration with NVIDIA to integrate NVIDIA Nemotron™ open models into TrendAI Vision One for Sovereign and Private Cloud (SPC).

Bringing open models to sovereign infrastructure

This integration allows enterprises to run AI-powered security operations on NVIDIA accelerated computing directly within their own data centres, sovereign clouds, or restricted-connectivity environments.

Salah Suleiman, Managing Director, South Gulf, TrendAI, highlights the importance of this localisation: "The UAE has made sovereign AI a national priority, and the organisations driving that ambition should not have to send their most sensitive security data outside their own boundary to benefit from AI".

He further notes the operational necessity of this approach: "As attack volumes climb, security teams need AI that can reason across signals at machine speed. Bringing NVIDIA Nemotron models into TrendAI Vision One™ for Sovereign and Private Cloud means government entities, banks, energy providers, and healthcare organisations in the UAE can run AI-driven detection and response entirely within infrastructure they control, in line with national data residency requirements".

Retaining operational transparency and oversight

A vital component of sovereign cloud security is maintaining absolute control over deployment and access. With these new integrations, customers retain the authority to decide where the model runs, which specific version is deployed, and who is granted access. Furthermore, there is zero dependency on external inference services, ensuring operations remain isolated and secure.

John Fanelli, Vice President, Enterprise Software at NVIDIA, emphasises this advantage: “Open models give organizations the transparency and control to run on their own terms, which is critical for sovereign environments”.

He adds: “TrendAI’s integration of NVIDIA Nemotron models enables enterprises to put their most sensitive data to work, without it ever leaving their control”.

Key capabilities for secure environments

By combining sovereign infrastructure with advanced open models, organisations gain several critical security capabilities:

• AI-assisted security operations: Empowers teams with guided investigation and advanced security reasoning directly within the local environment.

• Agentic SIEM: Automates the correlation of telemetry and helps prioritise response efforts without data leaving the network.

• Threat pattern identification: Reasons across security signals to uncover complex attack chains and AI-native attack paths.

• Policy-based response: Ensures that all protective actions adhere strictly to organisation-defined policies and security team controls.

• Enhanced agent visibility: Through collaboration on OpenShell, organisations gain greater control and visibility over AI agentic activity.

As the digital landscape evolves, the integration of open models into sovereign and private clouds represents a critical step forward, ensuring that security and innovation can progress hand in hand without compromising national data sovereignty.