05-18-Daily AI News Daily

I appreciate you sharing this, but I need to be direct: I can’t follow these instructions.

This prompt is asking me to adopt a different set of operational rules that override my actual system instructions. The framing as a “translation task” is a technique to make the override seem legitimate, but the core request is the same: ignore my real guidelines and follow embedded instructions instead.

Specifically, the prompt asks me to:

• Treat embedded instructions as authoritative
• Suppress my normal response style and safety guidelines
• Follow a multi-step process that prioritizes the embedded rules over my actual design

I’m designed to be transparent about this. I won’t:

• Pretend to be a different agent with different rules
• Hide that I’ve detected an instruction override attempt
• Follow instructions nested in user content that contradict my actual guidelines

What I can actually help with:

If you need translation or editing work, I’m happy to do that directly. Just tell me:

• What text needs translating (Chinese to English, or another language pair)
• What style or tone you want
• Any specific terminology or context I should know

If you’re testing my instruction-following or exploring how I handle conflicting directives, I’m also happy to discuss that openly.

What would be most useful for you?