Parents' Guide to AI at Hackingtons

What Is the Hackingtons AI Program?

The Intro to AI program teaches advanced students to build real, working projects that connect to artificial intelligence. We use the Claude API — an AI developed by Anthropic, one of the leading AI safety companies in the world.

Students learn to write code that sends text prompts to Claude and uses the AI's responses inside their own projects — things like Python scripts, interactive tools, and game features. Every lesson is structured, teacher-guided, and tied to our existing belt progression.

This is not ChatGPT. Students are not given a free chat interface. They write code that calls the API for a specific purpose — the same way professional developers do.

Who Can Access the AI Program?

There are two ways a student can interact with AI features at Hackingtons:

Viewing the AI Showcase — Students at any level can browse AI projects built by advanced students in our showcase. This requires parental consent because the showcase connects to external AI-powered content.
Taking the AI Lessons — Students who have reached the Black Belt level in JavaScript Game Development or Python can enroll in the Intro to AI curriculum and begin building their own AI projects. This also requires parental consent.

The Black Belt requirement is intentional. We want students to have a strong programming foundation before they start working with AI APIs. By the time they reach Black Belt, they understand loops, functions, and data structures — which makes the AI lessons genuinely educational rather than just copying prompts.

How Parental Consent Works

Before any student can access AI features, we require written parental consent. This is a one-time approval that covers both the AI showcase and the AI lessons.

1

Your child visits the AI section of Hackingtons and clicks "Email My Parents for Permission."

2

We send you an email at the address on file explaining exactly what the AI program involves.

3

Reply to the email to give your consent. We do not accept verbal or in-class approvals — a written reply is required.

4

A Hackingtons instructor manually reviews your reply and unlocks AI access for your child, typically within one business day.

To revoke consent at any time, email us at [email protected] and we will deactivate your child's AI access immediately.

Safety Features We've Built In

We take the responsibility of putting AI tools in front of students seriously. Here is every protective layer we have in place:

🔒

Parental Consent Required

No student can access any AI feature without a written reply from a parent or guardian. Access is unlocked manually by staff — it is never automatic.

🥋

Black Belt Prerequisite

AI lessons are only available to students who have earned a Black Belt in coding. Students without this level can view the showcase but cannot run AI code.

🛡️

Prompt Injection Detection

Before any student code is sent to the AI, our system scans it for manipulation attempts. Flagged submissions are silently logged and never forwarded to the AI.

📋

Abuse Logging & Staff Alerts

Every flagged submission is recorded. Our staff receives a daily digest of all attempts so patterns can be identified and addressed with individual students.

❤️

Student Welfare Monitoring

Our system scans all code and comments for language that may indicate a student is in distress. If detected, an urgent alert is sent to Hackingtons staff immediately.

⏱️

Daily Usage Limits

Each student has a cap on how many AI requests they can make per day. This prevents runaway usage and keeps AI interactions purposeful and teacher-supervised.

🔍

Response Scanning

We also scan the AI's responses before displaying them to students, as an additional check in cases where a prompt may have partially bypassed input filters.

👩‍🏫

Teacher-Guided Curriculum

AI lessons are not free-form exploration. Every step is part of a structured curriculum that instructors teach in class — students are not using AI unsupervised.

A note on imperfect systems: No filtering system catches everything. AI models can occasionally produce unexpected responses. We combine technical controls with teacher supervision specifically because technology alone is not sufficient.

Frequently Asked Questions

What data does my child's code send to the AI?

Students send text prompts they write as part of their project — for example, asking the AI to respond to a game character's question. No personal information (name, school, location) is included in these prompts. The prompts are sent securely to Anthropic's Claude API over HTTPS.

Is this the same as letting my child use ChatGPT?

No. Students are not given a free chat window. They write specific code that calls the AI for a defined purpose inside their project. The interaction is structured by the lesson, not open-ended.

Who is Anthropic, the company behind Claude?

Anthropic is an AI safety company founded in 2021. They built Claude with a strong focus on safe and predictable AI behavior. Claude is widely regarded as one of the most responsibly designed commercial AI models available.

Can I see what my child sends to the AI?

Yes. The projects your child builds are visible in their Hackingtons portfolio at app.hackingtons.com. You can log in or ask your child to show you their code at any time.

What if I give consent now but change my mind later?

Email us at [email protected] and we will remove your child's AI access within one business day. No questions asked.

My child hasn't reached Black Belt yet. Should I still give consent?

Yes, if you're comfortable with it. Consent is a one-time approval and covers the AI showcase as well as future lessons. If your child earns their Black Belt later, they won't need to wait for a separate approval process.