CIACC AI Presentation 🤖

Poll

How confident do you feel guiding youth AI use right now?

• Very confident

• Somewhat confident

• Not confident yet

• I’m here to learn basics

Word Cloud

When you hear ‘AI + youth wellbeing,’ what’s your biggest concern?

Multiple Choice

Generative AI is best described as…

• A search engine that finds facts

• A tool that predicts likely answers based on patterns

• A program that always cites its sources

• A tool that can’t make mistakes

Categorize

Options (10)

A student says they want to hurt themselves. Ask AI what to do next.

Paste a counselor note or SBYS case summary and ask AI to suggest interventions.

Turn a rough staff note into a professional summary, removing any names or unique details.

Role-play a chat script for a conflict scenario with generic roles (Student A/Student B).

Create conversation starters for a student returning after a suspension (no case details).

Draft a parent email about attendance concerns using a made-up student name (no identifiers).

Write a short flyer about healthy sleep habits for teens (general tips only).

Create 10 icebreaker questions for a middle school small group on stress management.

Translate a generic family reminder message into Spanish and Portuguese (no names).

1. Summarize a public article on teen mental health and social media into 5 key takeaways.

Sort each scenario: Green / Yellow / Red using privacy, safety, boundaries.

🟢 GREEN

🟡 YELLOW

🔴 RED

Match

Match the verification step to the reason it matters.

• Cross-check with a trusted source

• Scan for bias

• School-appropriate language

• Pause on sensitive info

• Ask for sources

reduces misinformation

prevents harm/stereotypes

student-safe delivery

move to established supports

transparency/verification trail

• reduces misinformation

• prevents harm/stereotypes

• student-safe delivery

• move to established supports

• transparency/verification trail

Open Ended

Write one AI prompt you could use in your role this month. Your prompt should include Goal, Format, Guardrails, Context.

Guardrails to consider:
Do not include student names, identifying details, or case information. Do not request crisis guidance, safety planning, diagnosis language, or risk ranking.

Examples of appropriate tasks (choose one or your own):

• Draft a family newsletter blurb

• Translate a program announcement

• Create a small-group discussion guide

• Write a restorative conversation script (generic roles)

• Generate an agenda/checklist for an event

Open Ended

In 2–3 sentences, reflect on what you’ll take into your work this month:
1) One guardrail you will apply every time you use AI is: ______.
2) One “Yellow zone” situation you’ll slow down on and verify with a person is: ______.
3) One safe use case you’re ready to try is: ______.