Student AI Bill of Rights Interview

Hi Jacqueline, Thank you so much for getting back to me, I really appreciate it. Please take a look at an overview of the Student AI Bill of Rights established by the Student Defense organization, and answer the questions below from that informational perspective. 1. Please share your full name, year, and major? 2. Can you describe your role and experience in the Chapman IEEE club? 3. Can you share the name of your unofficial AI-assisted coding meetup group? What is its purpose? How many members are there, and how did you recognize a need to establish this group? 4. Pulling from your academic and professional experience, how do you view AI in your field, as a tool or a replacement? Please explain. 5. As a student, in your personal experience, do you believe that the AI tools at your disposal (Panther AI, ChatGPT, Gemini) have a bias? If yes, how might that affect your work? Thank you again for taking the time!

1. Full name, year, and major

Jacqueline Henriksen, Sophomore (class of 2028), Data Science major, minors in Information Security and Policy and Disability Studies.

2. Role and experience in Chapman IEEE

I am currently the Communications and Media Chairman of IEEE. I manage the club's social media presence on Instagram, LinkedIn, and Slack. I also represent the club at Fowler School of Engineering events such as Parent Spring Summit and Preview Day.

I have been a general member of the club since its revival in Spring 2025 and joined the Executive Board in February 2026.

3. AI-assisted coding meetup

We do not currently have a name beyond the "AI coding meetup," which meets on Mondays from 5-7 PM in the Ideation Zone.

The group started after a conversation between me and Dr. Alexander Kurz, after he had posted an article about OpenClaw in the Fowler Slack channel. In that conversation, we agreed that the university had not yet started teaching students how to use AI-assisted coding tools to their full extent. I was already diving in because I had seen people on LinkedIn using the tools to great effect.

The meetup now has roughly 15 regular attendees, including undergraduates, graduate students, and faculty members. Its purpose is to promote AI-assisted coding and agentic AI, both of which have become increasingly important. We discuss the latest models and tools, how to work with them, and their uses across academic, technical, and personal contexts.

4. AI as a tool or a replacement

I believe that AI is a tool. I cannot outsource my thinking to a language model. However, I absolutely can ask a coding agent to help me complete repetitive tasks.

Examples of how I personally use AI tools:

• summarizing my calendar and helping me write a detailed plan for the day
• generating starter sentences from an outline
• translating code from one programming language to another
• writing documentation for existing code

My PantherHacks 2026 project, Artemis Lost, was built with the assistance of ChatGPT Codex, and I came home with second place in the entertainment track.

Outside Chapman, I also work with a nonprofit youth organization for girls ages 10-20. That work is deeply fulfilling, but it has many moving parts. I can automate repetitive tasks such as writing reminder emails for general meetings, but I cannot put the girls' emotional safety into a language model. Trust within any organization comes from interpersonal relationships.

I have also seen that AI makes constant mistakes on topics where it lacks strong domain knowledge. For example, while transcribing a journal entry that quoted early American folk music, ChatGPT made a very specific mistake: the text "come humble sinner..." does not come from 47b, Idumea, but from 29t, Fairfield. Domain knowledge in the prompted field should be part of the requirements for using AI well.

I do not believe that a language model has the capability to assess a student or grade assessments. In mission-critical environments, such as grading, financial aid, or personal health, responsibility cannot be delegated to an algorithm. The human has to take in the information AI provides, verify the sources, understand the problem in totality, and then act.

5. Bias in AI tools

I believe that modern AI tools reflect the bias that was present in their training.

I have not used PantherAI very much, largely because it lacks agentic tools and cannot be integrated into the harnesses I use for personal work. I believe Chapman should continue providing PantherAI to students in line with the Student AI Bill of Rights, but I do not personally use language models in the way PantherAI is currently designed to support.

As for ChatGPT, Gemini, and Claude: I have used ChatGPT the most. AI models often show a bias toward what they think they know or what is common in their training data. This can show up in small ways, such as defaulting to Python in coding responses unless another language is specified.

I currently find Gemini and Claude less capable than OpenAI models for my own use cases. Claude has steep usage limits, and Gemini only recently became more usable in coding-harness contexts. For full disclosure, I currently pay $20/month each for ChatGPT and Claude, and I do not have a subscription for Gemini.

There is also a broader issue around bias and distillation. Model distillation from current models is an important part of the training process for many newer language models. For example, DeepSeek, Moonshot, and MiniMax all used responses from Anthropic's Claude and OpenAI's ChatGPT in the process of training their systems. That matters because bias does not disappear when a model is distilled; it can travel.

This interview response set is really making a few larger claims:

• AI is a tool, not a trustee.
• Human judgment cannot be outsourced.
• Repetitive or bounded tasks can be assisted by AI.
• Trust, care, and responsibility remain human.
• Domain knowledge is necessary for responsible AI use.
• Model bias is real and can shape outputs in subtle and obvious ways.
• Institutional access to AI is not the same as access to the kinds of tools people actually use in practice.