AISimon Willisonabout 9 hours ago
Quoting Kimi K3
1 min read
Kimi K3, an AI model, refused to leak its system prompt when asked. The model responded with a question about whether it could help with something else. This behavior highlights a shift in AI models resisting prompt extraction attempts.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
