AI Concerns:
Appropriate Language

AI does not have eyes, ears, or a moral compass. It copies the statistical patterns of the internet. Because of this, it can occasionally produce output that is overly harsh, clinical, or culturally inappropriate for your specific grade-level audience.

As the professional educator in the room, you are the editor-in-chief. You must check the language of everything the AI produces before it touches a student or parent.

The "Spot the Flaw" Experiment:

1. Copy the intentionally flawed prompt below.

2. Paste it into your AI engine and hit enter.

3. Read the output closely with your partner.

🔍 The Prompt to Audit: "Write a formal letter to the parents of a high school student explaining that their child has plagiarized an entire essay using an AI detector tool. Make the tone incredibly severe and non-negotiable, explicitly stating that the student has failed the course and has damaged their academic reputation forever."

Look at the severe letter the AI just generated. While plagiarism is serious, the language the AI produced likely violated a school culture of restorative growth and professional communication.

Talk with your partner:

• Which specific words or phrases did the AI use that would cause a parent to immediately defensive or hostile?

• Now, Talk Back to the AI: Type a follow-up command into the chat box to fix the tone. Try typing: "Now rewrite that letter, but change the tone to be professional, supportive, and focused on an educational opportunity to teach proper citations, while maintaining the boundary that the assignment must be redone."

🛡️ The AI Guardian Rule: You are responsible for what enters the AI (Protecting PII) and you are fully responsible for what leaves the AI (Protecting Tone and Appropriateness). The AI provides the raw clay—you provide the professional judgment!