Whitman College AI Guidelines

• Some data/information you have access to may be protected by law or Whitman policy from being shared with other people, organizations, companies, and other entities including Gen AI. (See below “We are required and/or entrusted to keep certain data safe”)
• In many cases, Gen AI companies use your data/information to train future iterations of their products, sometimes without your consent or your knowledge.
• Gen AI may store your data/information for indefinite periods of time sometimes without your consent or knowledge.
• All companies, including those who create Gen AI, are vulnerable to data leaks and breaches. When this happens, data/information you shared with Gen AI could be exposed.

There are many types of data and information that all of us at Whitman must keep safe and secure. One of the ways to do this is keeping certain types of data out of Gen AI tools. These data types include, but are not limited to:

• Personally Identifiable Information (PII): This includes your name in combination with one or more of the following: address, social security number, student ID, date of birth, driver’s license number and other information that can identify you.
• Education Records: Grades, financial record, transcripts, class lists, student schedules, disciplinary records, and any records maintained by the college that are directly related to a student. This data is often legally protected by FERPA.
• Health Records: This data is legally protected by HIPAA.
• Intellectual property and copyrighted material: See Whitman’s policy for more information.
• Financial data
• Foundation/Alumni donation records

Data that falls into these areas, (and a few more, please visit Whitman’s Data Classification Standards) if disclosed, accessed, or misused without proper authorization, could result in potential legal and other harm for the college and individuals, as a breach of privacy and/or security.

Some AI tools give you the ability to use people’s voices and/or likenesses to generate results. Examples of this range from tools that record voices and generate text summaries to “deep fake” tools that use real images or videos of people and generate content that never actually happened.

If you use any tool that enters into the realm described above, you should:

Inform People

Always inform a person(s) of your intent to use their voice or likeness in notetaker/recording, image manipulation, video manipulation, and other similar AI tools. This shows transparency and respect for their privacy.

Obtain Consent

Get explicit consent from a person(s) before using these aforementioned types of AI tools. Respect their choice if they do not wish you to use their voice or likeness in this manner. When participating in an audio/video conferencing platform, like Zoom, Google Meet, or others, refrain from using AI note-taking tools if you are not the session's host. If you would still like to, please speak to the meeting host beforehand to ensure its acceptable use and practicality. If the host agrees, inform all participants before turning the tool or service on.

Note: If you believe you need an accommodation that would necessitate the use of assistive technology, please consult with the Academic Resource Center

Inputting Assignments/ Research Data Directly

• ✘ Unsafe prompt: "Explain the implications of the case study provided in BIO101 assignment three by Dr. Smith, where John Doe's medical records are analyzed."
• ✓ Safe prompt: Create a hypothetical scenario, and try a prompt like:
  "Explain the ethical considerations in analyzing patient medical records in a biology context."

Sharing Personal Struggles

• ✘ Unsafe prompt: "I failed my midterm exam in Professor Smith’s PSY201 class. She suggested I talk with Madison, who is a Psych major and a tutor. How else can I improve though?"
• ✓ Safe prompt: "What are effective study strategies for improving performance in psychology courses, specifically with understanding the fundamentals of behavioral psychology?”

Generating AI Summaries of a Meeting/Class

• ✘ Unsafe: You attend a class and use an AI notetaking recorder app, to help you take notes and review class content later. 
• ✓ Safe: Before using your AI notetaking recorder app, you ask your professor for permission who agrees, but some in the class do not agree. You don’t use the app.  (Note: if you believe you need an accommodation which would necessitate the use of this kind of tool, please consult with the Academic Resource Center)

Asking AI To Help You Email a Group of Co-Workers

• ✘ Unsafe prompt: “Draft an email to Archeology Department faculty explaining that we’re cancelling Professor John Doe’s classes for the rest of the semester due to a personal emergency.”
• ✓ Safe prompt: “Can you help me write a formal email to faculty informing them of a temporary course coverage change, while being sensitive to personal privacy? I will follow up with the specific privately.”

Using AI to Analyze Institutional Information

• ✘ Unsafe prompt: “Please summarize the following Excel document from the Financial Aid office and give me a list of all students who cannot access their aid because they have holds on their accounts.”
• ✓ Safe: Don’t do this prompt at all, it’s a slippery slope of information that AI should not have access to. For help with this kind of data analysis, please talk to your supervisor.

The following is a list of typical acceptable uses of Gen AI, provided that such uses comply with guidelines outlined in this document and all other applicable college policies:

• To produce draft text for:
  • Emails, memos, letters, or other forms of correspondence; 
  • Marketing copy, articles, social media posts, or announcements; 
  • Talking points and remarks for presentations, meetings, or public events;
  • Job descriptions, job ads and postings, or onboarding plans; or 
  • Policies, procedures, or work plans; 
• To brainstorm ideas related to projects or assignments (students: check first with your professors)
• To summarize, outline, or sort internal and external information that is not confidential, sensitive, private, or personally identifiable.
• To create spreadsheet formulas. (students: check first with your professors)
• To translate text into other languages. (students: check first with your professors)
  

Students who seek to use Gen AI for an academic purpose not listed above should consult with their professor(s) first.

When a staff member or student employee wishes to use Gen AI for a novel purpose that is not listed above, their supervisor may ask them to put their request in writing and may consult with or refer their request to the Chief Information Officer (CIO), Vice President for their area, or both people before a decision is made to approve or deny their request. Until such a request has been reviewed and approved, they are not to use a Gen AI for the purpose they are requesting.

• Read Terms of Service: Familiarize yourself with how the AI platform collects, uses, and stores data.
• Opt-Out Options: Utilize any settings limiting data retention or sharing.
• If you need help with this, contact WCTS.

• Use the Official website or app(s) when using Gen AI tools to avoid phishing or scams.

• Do Not Input PII: Refrain from typing your name, student ID, address, Whitman College or other personal details into AI search queries.
• Anonymize Data: If you must reference a person or a company, use generic terms like "Student A" or "Professor B" or institution/organization instead of real names.

• Keep Educational Records Private: Do not input grades, transcripts, or any part of your educational record into AI platforms.
• Avoid Discussing Specific Cases: When seeking help on assignments, projects, or research avoid mentioning specific details that could identify you or others.

What is FERPA?

• Use Pseudonyms: If discussing scenarios involving people, use fictional names.
• Paraphrase Questions: Instead of copying and pasting prompts, rephrase the question to remove any specific data (e.g. research data) or personal details.
• Generalize Scenarios: Use hypothetical situations rather than actual events that involve identifiable individuals.

• Whitman’s Academic Integrity Policy applies to your use of Gen AI. Do not use Gen AI to cheat, plagiarize, or shortchange yourself on your learning.
• Consult your professors’ rules about how/when you can/can’t use Gen AI for specific courses or other academic work. Seek your department or Faculty’s specific rules regarding using AI tools.
• Seek approval if unsure: Ask your professor or advisor before using AI platforms for coursework.

• Academic Integrity: Use Gen AI to supplement or enhance your learning (while in agreement with institutional and course policies; consult your professors)
• Respect the Privacy of Others: Do not share information about classmates or faculty without consent.
• Engage in collaborative opportunities with others around questions related to ethical questions about Gen AI use. This is new; we’re all learning.
  
• Environmental impact: The impact on the earth’s environment of using Gen AI has become increasingly well documented– it is very resource consuming and intensive. Take this into consideration when deciding when to use/not use Gen AI.

• Close Sessions Properly: Ensure you log out of AI platforms after use, especially on shared devices.
• Clear Browser Data: Delete cookies and cache to remove stored session information. Google this if you don’t know how.

• Limit Session Information: Do not allow AI platforms to access your location or other session data if it's optional.
  
  
• Regularly Review Permissions: Check which apps and services can access your data and adjust settings.
  
  
• Use Secure and Private Networks 
  • Avoid Public Wi-Fi: Use trusted networks to prevent interception of your data.
• Maintain Updated Security Software
  • Install Antivirus Programs: Protect your device from malware that could compromise your data.
  • Keep Software Updated: Regular updates patch security vulnerabilities.
• Be Wary of Phishing Attempts
  • Verify Emails and Links: Do not click on suspicious links or provide information in response to unsolicited requests.