TechNewsReel
Live

NVIDIA Leads 37-Company Alliance to Secure Open-Weight AI Models

Tech giants including Microsoft, IBM, and Red Hat unite to build open-source defenses for the AI stack, arguing transparency beats proprietary lock-in.

TechNewsReel Newsroom · July 27, 2026

A New Defense Coalition

NVIDIA has announced the Open Secure AI Alliance, a coalition of approximately 37 inaugural partners committed to developing open tools for securing the AI stack. The alliance includes Adobe, Capital One, Cisco, Cloudflare, CrowdStrike, Databricks, Dell Technologies, HPE, Hugging Face, IBM, Microsoft, Palantir, Palo Alto Networks, Red Hat, Salesforce, SAP, ServiceNow, Siemens, and Snowflake.

Open Models as Security Assets

The alliance's mission is to safeguard everything from open-weight models to agent harnesses through transparent, community-driven security tools. Rather than treating open-source AI models as vulnerabilities, the coalition argues they increase visibility for security practitioners and broaden defensive capabilities across the industry.

The initiative emerges as AI systems become embedded in critical infrastructure spanning banking, healthcare, and telecommunications. Security failures in these models now represent systemic risks beyond individual companies.

Rejecting Proprietary Lock-In

"Trying to secure this landscape by fragmenting behind closed, proprietary boundaries simply shifts risk rather than solving it," said Chris Wright, CTO and SVP of Global Engineering at Red Hat.

The alliance challenges the notion that AI security requires proprietary walls. Instead, the group advocates for standardized, open-source defense layers as accessible as the models they protect.

Industry Context

The announcement arrives amid ongoing policy debates over open-weight AI models. A separate coalition including Microsoft and Meta recently urged US policymakers to avoid broad restrictions on open-weight models, warning that excessive regulation could stifle innovation.

Preventing a Security Monopoly

By creating shared defensive tools, the alliance aims to prevent what members describe as a "security monopoly" by proprietary AI labs. The goal is ensuring organizations defending against AI-powered cyber threats have access to the same tooling and intelligence as those building the models.

The coalition's composition spans chipmakers, cloud providers, enterprise software vendors, and AI research organizations, reflecting the breadth of stakeholders invested in securing the AI supply chain.

Sources

Get a notification when a big story breaks. A few a day at most — no spam.