AI Leaders Propose SAFE Guidelines for Cybersecurity Transparency
Covered by 2 sources
Read full postThe Open Secure AI Alliance, including NVIDIA, Cisco, CrowdStrike, Hugging Face, and Red Hat, has proposed the SAFE guidelines to enhance cybersecurity transparency for agentic AI systems by sharing incident data and best practices. These guidelines aim to foster collective defense through confidential incident sharing and evidence-based recommendations, complementing existing open-source AI security tools.
Covered by 2 sources
- Nvidia is quietly staffing a new AI safety team as it doubles down on open models· Business Insider
- Nvidia’s Open Source Alliance Is Missing Some Key Names: OpenAI and Anthropic· Wired
- A Chinese model saved Hugging Face. It is not in Nvidia’s new alliance.· The Next Web
- Industry Leaders Unite in Open Secure AI Alliance for AI Safety and Security· 2 sources
- Nvidia, Palantir, Hugging Face join 34 others in race to defend open-weight AI from cyber threats· 2 sources
