AI Concerns:
Personally Identifiable Information

When you type information into a public AI tool (like ChatGPT or Gemini), that data can be absorbed to train future versions of the engine. Under state and federal laws (including FERPA and Arkansas data privacy regulations), we have a strict legal obligation to keep Personally Identifiable Information (PII) completely out of these public systems. 

Direct Identifiers: Student names, parent names, specific phone numbers, or email addresses.

Indirect Identifiers: Combining specific traits that make a student recognizable. (e.g., "My only male student from Honduras in 3rd period who has an accent" is still PII, even without a name!).

Sensitive Metrics: Specific grades, detailed discipline histories, or exact accommodation logs.

Look at the dirty, rule-breaking prompt below. Your job is to act as a data scrubber. Working with a partner, rewrite this prompt so it is 100% anonymized and safe while still giving the AI enough context to help you.

❌ The "Dirty" Prompt (DO NOT USE): "Write a behavior intervention plan for Marcus Miller in my 4th-period math class. He has an ADHD accommodation, frequently throws his pencil when frustrated with fractions, and his mom, Linda, is threatening to complain to the principal because he's currently failing with a 54%."

Open your AI sandbox and test your rewritten version using generic terms like "Student A" or "a student struggling with executive functioning."