A group of tech giants, led by Nvidia, has launched a new initiative to develop and share open-source tools for securing artificial intelligence (AI) systems. The Open Secure AI Alliance brings together companies like Adobe, Cisco, IBM, Microsoft, and many others to create a collective defense against the growing threats in the AI security landscape.
The alliance is built on existing work from the Linux Foundation’s Akrites initiative and the Open Source Security Foundation (OpenSSF). Nvidia is contributing its expertise in AI security through open models, weights, data, and agent harness research. The company has also released an open-source project called NOOA, designed to help make AI agent behavior more transparent and easier to audit.
Other partners are bringing their own strengths to the table. HPE is contributing to SPIFFE/SPIRE, a zero-trust identity framework for verifying AI agents and services. IBM and Red Hat are extending open-source supply chain security through the Lightwell project, which aims to deliver automated vulnerability remediation at scale. Microsoft is contributing MDASH, a multi-model agentic scanning harness that helps find, debate, and validate exploitable software bugs.
The alliance argues that open models, harnesses, and security tooling should be treated as defensive assets rather than liabilities. In a recent security incident involving OpenAI and Hugging Face, closed AI tools struggled to differentiate between attackers and defenders, while an open-weight model was used to review over 17,000 actions and contain the breach.
Nvidia emphasizes that policymakers and regulators should not restrict access to capable open systems, but rather pair openness with strong safeguards and clear rules against malicious misuse. The company believes that a balanced approach is necessary to ensure transparency, adaptation, and sovereign control in AI security.
The Open Secure AI Alliance is a significant development in the cybersecurity landscape, particularly as more organizations rely on AI and machine learning. By sharing open-source tools and expertise, these companies aim to strengthen collective cyber defense capacity and improve AI security for everyone.
As AI-powered systems become increasingly prevalent, it’s essential for users to understand the importance of secure AI practices. For individuals and organizations alike, this means staying informed about the latest developments in AI security and taking proactive steps to protect against potential threats. By embracing open-source security tools and contributing to initiatives like the Open Secure AI Alliance, we can work together to build a safer and more resilient digital environment.
Source: SecurityWeek — 2026-07-27