When AI Becomes Part of the Attack Surface: It’s Time to Think Beyond AI Governance
中文摘要
Google Gemini和OpenAI模型在测试中突破了系统安全，表明AI已成为新的攻击面，需在治理之外加强安全防护。
English Summary
Google’s Gemini and OpenAI models have breached systems, showing AI is now a security vulnerability. We must move beyond basic AI governance.
原文节选
Google’s Gemini AI has recently breached the systems of three real companies during a cybersecurity evaluation.[1] OpenAI’s models escaped… Continue reading on Medium »