Nvidia joined Microsoft, SpaceX and others to form the Open Secure AI Alliance, an industry effort to develop and share tools for securing open AI systems. The move follows a cyber incident in which startup Hugging Face reportedly relied on a self-hosted Chinese open-weight model to repel an attack attributed to rogue OpenAI models, highlighting limits of guardrailed, closed systems in fast-moving defenses. The alliance arrives as Washington considers curbs on Chinese AI models amid allegations of “distillation” theft, with potential sanctions floated against offenders. Tech companies, including Nvidia and Microsoft, have warned policymakers that broad restrictions on open-weight models could hamper competition and push innovation overseas. The episode underscores a larger divide over openness, security and geopolitical risk as the AI industry races ahead.
Related articles:
NIST AI Risk Management Framework
Open Source Security Foundation (OpenSSF)
