Beyond the Model: Context, Harnesses, and the New AI Security Boundary
中文摘要
AI安全重心正从针对模型的攻击转向保护上下文与套件，这构成了全新的安全边界。
English Summary
AI security is shifting focus from model-centric threats like prompt injection to securing the context and harnesses that define the new security boundary.
Original Excerpt
For the past few years, most AI security conversations have focused on the model. Teams worried about prompt injection, jailbreaks… Continue reading on Medium »