# Appendix C: AI Use and Disclosure

This appendix restates the course AI policy from the syllabus. **When in doubt, disclose.**

## The General Principle

In this course, students **learn to use and collaborate with AI tools**. Their use is permitted on all assignments except as noted below.

Because learning about AI is the very purpose of this course, you are expected to engage with AI tools **thoughtfully, honestly, and in accordance with the guidelines below**.

> AI tools may be used on select assignments, as instructed, **provided their use is disclosed and properly cited**. Undisclosed use on any assignment where it is permitted — or any use where it is restricted — constitutes a violation of academic integrity and will be treated accordingly.

## Assignment-Specific Guidelines

| Assignment Type | AI Use | Requirements |
| :--- | :--- | :--- |
| **Homework assignments** | Permitted with disclosure | Cite which tool was used and describe how |
| **Labs / hands-on activities** | **Required** | AI tools are a core part of lab activities |
| **Quizzes** | Not permitted | All work must be your own |
| **Midterm exam** | Not permitted | Closed-book, no AI tools of any kind |
| **Capstone project** | Permitted with disclosure | Final analysis, arguments, and conclusions must reflect your own thinking; AI may assist with **research and drafting only** |

```{note}
Note the second row. On the **⚙️ Hands-On** activities throughout this book, using an AI assistant is not merely allowed — it is **expected**. Asking an assistant to explain a confusion matrix, interpret a loss curve, or unpack an error message is part of the exercise. That is a different skill from asking it to do the work, and this course is trying to build the first one.
```

## Disclosure Format

When AI use is permitted and you have used it, include this at the end of your submission:

> **AI Disclosure:** I used [Tool Name] to [brief description of how it was used — e.g., brainstorm ideas, check grammar, summarize a source]. The final responses and conclusions are my own.

### Examples

> **AI Disclosure:** I used Claude to explain the difference between precision and recall after I got confused by the Unit 3 confusion matrix. I wrote the analysis myself.

> **AI Disclosure:** I used ChatGPT to generate five candidate capstone topics and to check my draft for unclear sentences. Topic selection, data analysis, and all conclusions are my own.

> **AI Disclosure:** I used Gemini to interpret a `ValueError` in the Unit 5 code and to suggest what to check. The experiment design and interpretation of results are mine.

**Be specific.** "I used AI to help" is not a disclosure. Name the tool and name the task.

## Using AI Well in This Course

A practical distinction that maps onto how the policy is written:

| Productive use | Counterproductive use |
| :--- | :--- |
| "Explain what this error means" | "Fix my code" |
| "What am I missing in this argument?" | "Write my position paper" |
| "Give me three counterarguments to my thesis" | "What should my thesis be?" |
| "Is my explanation of overfitting accurate?" | "Explain overfitting for my homework" |
| "Quiz me on Unit 4 terminology" | "Answer these Unit 4 questions" |

The left column produces learning. The right column produces submitted text and no learning — and it is exactly what the exams and Q&A are designed to detect.

## A Note on Academic Integrity

This course is designed to help you develop **genuine AI fluency**, a skill requiring you to think critically, not just generate output. Submitting AI-generated work as your own without disclosure does not demonstrate learning — **it undermines it**. Students found in violation are subject to the institution's academic integrity procedures.

## Why This Policy Exists

AI tools are powerful, and knowing how to use them responsibly is itself a career skill (see Unit 8). This policy is not designed to restrict you. It is designed to ensure you graduate with **real knowledge and real capabilities that no AI tool can replace**.

```{important}
Consider the structure of this course. In Unit 5 you make a neural network fail. In Unit 9 you catch a data leak. In Unit 10 you build a discriminatory model and try to fix it. **An AI assistant can describe all of these to you. It cannot give you the experience of having done them** — and that experience is what will still be useful to you in twenty years, when every tool mentioned in this book has been replaced.
```

## University Policies

FGCU maintains central policies on recording class sessions, academic integrity and grievances, student accessibility services, academic disruption, and religious observances. Review them at:

[www.fgcu.edu/about/officeoftheprovost/coresyllabus-policy-statements](https://www.fgcu.edu/about/officeoftheprovost/coresyllabus-policy-statements)

## Submission Reminders

- **All assignments are submitted via Canvas.** Assignments sent by email, on paper, or by any other means will not be accepted.
- **Late work** requires written permission from the instructor obtained **before** the due date. Late penalties begin at 10% of the grade value and increase by 1% per day.
- If difficulty is affecting your academic performance, **talk to the professor as early as possible**. Office hours are M W 2:00–4:00 PM in ETI 112D.