Nvidia-backed Open Secure AI Alliance puts AI security and sovereignty in focus for Middle East

The Nvidia-backed Open Secure AI Alliance has launched an industry coalition to develop and share open tools, models and techniques for AI security, bringing together technology and cyber security companies including G42, Microsoft, IBM, Cisco, CrowdStrike, Cloudflare, Dell Technologies, Hugging Face, Palantir and the Linux Foundation. The initiative, announced in early August 2026, builds on the Linux Foundation’s Akrites work and the Open Source Security Foundation (OpenSSF) and aims to give cyber defenders access to AI systems they can inspect, modify and deploy on infrastructure they control.

"It could shift AI security from relying on proprietary solutions to adopting open, transparent and collaborative security tools," said Irene Corpuz, founding partner of Women in Cyber Security Middle East. "This would enable organisations to better audit, customise and validate AI systems while becoming more resilient through shared standards and community-driven innovation."

The alliance’s proponents say the open approach is intended to allow defenders to identify, remediate and disclose vulnerabilities in models and tooling, reducing dependence on a small number of vendors and avoiding single points of failure. Nvidia frames the move as complementary to proprietary platforms: open models and tools, the company argues, let organisations operate security capabilities on local infrastructure and adapt them to specific environments.

Abu Dhabi-based G42’s participation underscores the regional relevance of the initiative. Gulf countries including the UAE and Saudi Arabia have been investing in domestic AI infrastructure and pursuing data and technological sovereignty. For national agencies, critical infrastructure operators and regulated companies, the ability to deploy security models locally is central to meeting data residency requirements and aligning with national cyber security strategies.

• The alliance expands existing open-source security work from the Linux Foundation’s Akrites initiative and OpenSSF.
• Members include Nvidia, G42, Microsoft, IBM, Cisco, CrowdStrike, Cloudflare, Dell Technologies, Hugging Face, Palantir and the Linux Foundation.
• Saudi Arabia’s National Cyber Security Authority is consulting on AI Cyber Security Guidelines covering four areas: cyber security governance, cyber security defence, cyber resilience and third-party cyber security.

Open security presents trade-offs. Greater transparency can empower defenders to audit and customise systems, but it can also enable malicious actors to examine capabilities and identify weaknesses. Nvidia acknowledges that open models can be misused, including by weakening safeguards or adapting capabilities for cyber attacks, though it notes such risks are not exclusive to open systems. Corpuz cautioned that open does not equal inherently secure and stressed governance requirements: "The key is to implement strong governance, secure-by-design practices, continuous testing, access controls most especially, and threat monitoring to ensure the benefits outweigh the risks."

How the Open Secure AI Alliance will affect Middle Eastern adoption depends on its ability to translate collaboration into deployable tools and standards that align with sovereign infrastructure and local regulation. The Saudi Data and Artificial Intelligence Authority’s national strategy stresses responsible AI adoption, advanced digital infrastructure and cyber resilience — priorities that map directly onto the alliance’s objectives. If the initiative can produce verifiable, governable open defensive tools, Gulf organisations may gain options to retain strategic control over AI security while meeting national cyber and data-residency requirements.

Outlook

Progress will be measured by tangible outputs: shared model audits, vulnerability disclosure frameworks, secure-by-design toolkits and governance codes that member organisations adopt. For the UAE and Saudi Arabia, the alliance offers a potential mechanism to combine open security frameworks with sovereign deployment. But as industry and regulators weigh the benefits, the central challenge remains balancing openness with robust governance to prevent open AI from becoming an exploitable weapon as systems gain autonomous capabilities.