AI Usage at Wright State University (WSU)
On this page:
Get Started With Copilot
- What is Copilot?
- What is Artificial Intelligence (AI)?
- Acceptable Use of AI at Wright State
- Data Privacy and Security
- How to Disable Use of Data for Training
- Best Practices for Faculty and Staff
- AI Note-Taking Best Practices
- Acquiring a License and Procurement Considerations
- Additional Considerations
- Frequently Asked Questions
What is Copilot? 
Copilot is a secure, AI-powered conversational assistant. Copilot Chat is available to all students, faculty, and staff with an active Wright State account. Microsoft has several products called Copilot. Copilot Chat is a specific version provided by Wright State's Microsoft license. Other versions, such as Copilot for Microsoft 365 (M365), have different levels of access to content. For a full overview of Copilot Chat and the differences between it and Microsoft Copilot for M365, refer to https://learn.microsoft.com/en-us/copilot/overview
NOTE: Microsoft Copilot for M365 costs an additional $360, per license, per year. If employees wish to purchase a license, please complete this Purchasing Approval through your Fiscal Manager and CaTS.
What is Artificial Intelligence (AI)?
Artificial intelligence (AI) is a technology that allows computing devices to perform tasks that typically require human intelligence. AI can learn, understand, solve problems, and make decisions, and is at the forefront of emerging technology.
Generative AI includes, but is not limited to, tools like Copilot and ChatGPT, with built-in functionality such as AI meeting summarization that integrates with our environment.
On this page, you'll find more information about AI usage at Wright State as we begin to navigate this new territory.
Acceptable Use of AI at Wright State
As AI tools like ChatGPT, Copilot, and other generative applications become more integrated into academic and administrative activities, it's important to use them responsibly. This guidance outlines acceptable practices for using AI while protecting institutional, personal, and proprietary information. Additional updates may be provided as circumstances change.
Allowable Use
- Public or Published Data: Freely usable with AI tools.
- Controlled or Confidential Data: May be used only with AI tools managed by the institution and covered by contracts that:
- Explicitly protect institutional data.
- Prevent data from being used to train models.
- Ensure data is isolated from external access.
- Authorized AI Tools: University provided Copilot Chat and ChatGPT
- Acceptable Use Policy: All use must comply with WSU's institutional acceptable use policies
Data Classification Examples (non‑exhaustive):
- Public: published web content, course catalog text, public policies.
- Internal: routine business communications, non‑public drafts without sensitive data.
- Confidential/Controlled: student education records (FERPA), health information (HIPAA/PHI), payment card data (PCI), financial records (GLBA), research with confidentiality obligations, HR/personnel data, export‑controlled research (ITAR/EAR).
Minimum Controls by Classification:
- Public: allowed in approved or unapproved AI tools; still verify outputs.
- Internal: use only approved tools where feasible; avoid uploading attested sensitive details.
- Confidential/Controlled: only in WSU approved, contract‑bound tools with training‑off, auditing, and logging; apply least‑data‑necessary.
Student access: Students may use AI tools where permitted by WSU, their instructors, and consistent with Academic Integrity policies. Students should check with their instructors and their course syllabus regarding AI usage, should check the course syllabus and follow any guidelines set by the instructor. For additional information, refer to Policy 3710: Academic Integrity Standards and Process for Misconduct.
Prohibited Use
- Unauthorized AI Tools: Tools without institutional contracts or proper data-sharing controls must not be used with Controlled or Confidential data. This includes free or unmanaged versions of AI tools.
- Sensitive Information: Do not use unauthorized AI tools with student records (FERPA), health data, proprietary information, or other confidential data.
- Non-Public Output: AI tools must not be used to generate:
- Academic work not permitted by instructors
- Illegal or Fraudulent Activities: AI tools must not be used for any activity that violates laws or institutional policies.
Data Privacy and Security
Generative AI tools often process input data on external servers, raising the risk of data leakage.Sensitive, proprietary, or personally identifiable information (PII) entered into AI systems may be stored or reused without explicit knowledge. To protect data privacy and security, the following guidelines must be followed:
- Anonymize Data: Remove identifiable information from inputs to protect privacy
- Use Approved Tools: Only AI Platforms that have been vetted and approved by WSU's central IT team, CaTS, may integrate with WSU systems and data.
- Understand Data Retention Policies: Use platforms that have transparent data usage policies and do not retain input data without explicit consent
Report suspected disclosure or mishandling of data to the Information Security Office immediately by submitting the Security Incident Response Form.
How to Disable Use of Data for Training
Important Note: if/when using other non-University endorsed tools, refer to the specific Help section for that company on how to disable data training
1) Copilot Chat
When using Copilot with your WSU account, prompts and responses are not used to train foundation models. There’s nothing to “turn off.” Just be signed in with your WSU account and not a personal Microsoft account.
2) ChatGPT Free & ChatGPT Plus, ChatGPT Business (Team)
Free/Plus personal workspaces allow using your content to improve models unless you opt out. Chat GPT Business (Team) does not have the training feature turned on by default. To opt out of this feature for the Free and Plus versions:
A. Opt-out of training (keep history)
- Go to Profile > Settings > Data Controls.
- Turn Off: “Improve the model for everyone”.
– New chats won’t be used for training; your history still works.
B. Turn off chat history (also disables training for those chats)
- Profile > Settings > Data Controls.
- Turn Off: Chat history & training (wording may vary).
– New chats won’t appear in history and won’t be used for training.
Mobile path (Free/Plus)
- Open the ChatGPT app > menu > Profile > Settings > Data Controls.
- Turn Off “Improve the model for everyone” (or Chat history & training).
Best Practices for AI Usage
- Treat AI inputs as public: Avoid sharing personal, confidential, or proprietary information with AI tools. Any content/information input into the free AI tools will become part of the public domain for that resource.
- Protect sensitive data: Be cautious with data that could affect legal, regulatory, or contractual obligations.
- Verify AI outputs: AI responses may include errors, biases, or unverified content. Always review and use your own judgment.
Teaching & Learning Practices:
Research & Scholarship:
- Consult Research and Sponsored Programs (RSP) and sponsor terms before using AI to process human‑subjects or restricted research data.
- Document provenance and permissions for AI‑generated figures, code, and text in publications.
- Do not upload export-controlled or sponsor-restricted research data into any AI tool unless explicitly cleared by RSP.
AI Note-Taking Best Practices
The use of AI note-taking tools requires careful consideration of privacy, accuracy, and equitable access. Below are five areas to keep in mind when utilizing these note-taking services:
- Transparency
- Meeting Host Responsibilities
- Accuracy
- Accessibility
- Information Security and Storage
Transparency
To maintain transparency among all participants, clearly communicate the use of AI note-takers before enabling them in a virtual meeting. Make sure to specify the purpose of the note-taking tools, and identify the specific AI tool being used.
Meeting Host Responsibilities
The meeting organizer determines AI note-taking use, and therefore is responsible of informing all participants when these tools are being utilized. If objections arise from any participant(s), explore alternative note-taking methods such as manual recordings or application-based transcription options.
Accuracy
AI-generated notes must be reviewed for errors, verified, and formally approved by the represented group before being used for documentation purposes or widespread distribution. AI summaries should always remain objective, excluding irrelevant or private details. Make sure to address potential biases in AI-generated content to ensure fairness and accurate representation.
Accessibility
Ensure AI notes are accessible to all participants, considering disabilities and non-digital format preferences.
Information Security and Storage
Sensitive data must be handled in compliance with WSU data protection policies. Third-party AI apps may have different retention policies; therefore, meeting notes must be stored in approved WSU locations (e.g., network storage, SharePoint, or Microsoft OneDrive). For information on approve data storage locations for data, please see the Data Classification Matrix.
Acquiring a License and Procurement Considerations
When considering using and/or purchasing an AI tool, be mindful of the following:
- Supplier-hosted data requires extra caution due to higher risk.
Assume supplier tools may use Wright State data for training, so follow strict guidelines to manage and secure it effectively - Request that CaTS Information Security team review any new tools for be used for WSU business by completing the Cloud Service Approval process
ChatGPT Personal LicensesWSU does not currently have a contract with OpenAI and ChatGPT. However, students and employees have used the free version and purchased Plus and Business (Team) versions. We encourage students and employees to reference the below information to understand the different options. If employees wish to purchase Plus or Business versions, please contact your fiscal services manager. |
|||
|---|---|---|---|
| Tier | ChatGPT Free | ChatGPT Plus | ChatGPT Business (Team) |
| Price (USD/user/month) | Free | $20 | $25 - $30 |
| Where it runs | chatgpt.com (web), mobile apps | chatgpt.com + mobile apps | chatgpt.com workspace + mobile; shared projects & connectors |
| M365 app integration | N/A | N/A | N/A (uses connectors to Microsoft/Google tools) |
| Security/Compliance | Consumer; outside enterprise boundary | Consumer; not enterprise compliance | Workspace admin controls; Data not used to train OpenAI models |
| Notable limits | Lower message/tools limits; fewer advanced capabilities | Priority access to advanced models & features; higher limits than Free | Seat minimums (2+); admin console; flexible credits for advanced features |
| Who it's for | General/personal use | Power users/individual creators | Small/medium teams |
| How to Get | Self‑sign‑up | Self‑purchase | Upgrade to Business workspace |
| License Model | Individual (no contract) | Individual subscription | Self‑purchase |
| Typical Use | Personal Q&A, drafting, study help | Faster responses, advanced tools | Team collaboration, projects, connectors |
| Data Grounding | Web + files you upload in-session | Web + files you upload in-session | Files, connectors, and workspace GPTs |
| Access Priority | Standard | Priority vs Free | Higher limits than Free/Plus |
Additional Considerations
Legal Considerations
Important Note: Wright State University data must remain on servers located in the United States.
All AI tool usage must comply with applicable laws, regulations, and not use personally identifiable information (PII), which includes but is not limited to:
- Intellectual property rights
- Data privacy laws (non-exhaustive)
- Family Educational Rights and Privacy Act (FERPA)
- General Data Protection Regulation (GDPR)
- Health Insurance Portability and Accountability Act (HIPAA)
- Defense Federal Acquisition Regulation Supplement (DFARS)
- Export Administration Regulations (EAR)
- International Traffic in Arms Regulations (ITAR)
- National Institute of Standards and Technology (NIST)
- Payment Card Industry Data Security Standard (PCI DSS)
- Gramm-Leach-Bliley Act (GLBA)
- Copyright laws
- Contractual disputes involving AI-generated content
- Americans with Disabilities ACT (ADA) Considerations
- Content, particularly web and code, may not meet ADA requirements
- Mitigation of inaccurate AI-generated data
- Always validate output for correctness and be aware of potential bias
Ethical Considerations
It is vital to be mindful of ethical use of AI, which includes:
- Consent for AI Note Takers: Never use AI note takers without obtaining consent from all meeting participants.
- Data Sharing: Disclose the intention to share information gathered using AI tools.
- Misinformation Awareness: Be mindful of inaccuracies or biases in AI-generated data.
Fairness and Equity:
- Evaluate outputs for bias, stereotyping, and disparate impact; correct or avoid outputs that could cause harm.
- Ensure accommodations and equivalent alternatives when AI use could disadvantage participants with disabilities or limited access.
Frequently Asked Questions
Who decides how AI-generated notes are distributed?
- The meeting organizer is responsible for distribution, in consultation with participants.
- Notes containing confidential information must have restricted distribution.
Is AI use allowed in my course?
- It depends on the instructor and program policies. Follow your syllabus and departmental guidance. When allowed, cite AI assistance and verify accuracy. For more information, please refer to Policy 3710: Academic Integrity Standards and Process for Misconduct.
May I upload student records or HR documents to AI tools?
- Only in University‑approved tools (Copilot Chat and University-purchased versions of Copilot and ChatGPT) with proper contracts, retention settings, and access controls, and only when necessary. For University-purchased versions of ChatGPT Plus, the use of data for training must be disabled. Prefer anonymization.
How do I report a suspected AI‑related data incident?
- Contact the Information Security Office immediately and submit the Security Incident Response Form. Do not share additional data with the tool.
Where should I store AI outputs?
- Store in approved repositories aligned to data classification (e.g., secure network share, SharePoint/OneDrive with labels). Avoid local personal devices for sensitive content.
How much does an AI license cost?
- Copilot Chat is included in your WSU Microsoft 365 licensing. You do not have to do anything to enable or purchase it. Copilot for M365 is currently $360/user per year. ChatGPT Plus is $20/user per month. ChatGPT Business (Team) either has an annual cost of $25 per user/month, or a monthly cost of $30 per user/month. Please Note: This plan is for teams and requires 2 or more users
How long will my license be available?
- Copilot for Microsoft 365 licenses are renewed every fiscal year. Licenses purchased during the fiscal year will be prorated depending on when the purchase is made. ChatGPT Plus and Business (Team) is a monthly subscription and will be reviewed for renewal needs every XX period.
Can I use read.ai or similar note-taking applications?
- At present, external AI note-taking apps are not approved to integrate with Wright State's collaboration tools (M365 calendar, Teams and WebEx). Any tools approved in the future to supplement the built-in features of Teams and WebEx must be reviewed and approved by the CaTS Information Security team. 3rd Party add-ons to Microsoft 365 Teams and WebEx conferencing can be requested, but require a strong business case and CaTS approval. To request, please submit the Teams and Webex Third Party App Request.
Mandatory or Restricted AI Note-Taking
- AI note-taking should not be mandated due to consent considerations.
- AI note-taking is strongly discouraged for sensitive meetings such as performance evaluations, disciplinary matters, or discussions with students.
