AI Guidance

AI Utilization

As AI integrates into our daily research, teaching, and operations, we want to empower you to use these technologies safely and effectively. Our guidelines anchor on three pillars - ethics, privacy, and security. Check out the resources below to see which AI tools are approved for your specific department and get practical tips on how to confidently manage your data while using them.

AI Benefits and Challenges

Responsible AI use requires understanding both sides.

Benefits

      • Increased efficiency: Automates routine, time-consuming tasks so employees can focus on complex work 
      • Better decision-making: Analyzes large datasets to surface patterns and insights no human could process alone 
      • Improved service delivery: Faster, more personalized services - available 24/7, in multiple languages 
      • Enhanced research: Accelerates discovery, literature review, and hypothesis generation across disciplines 
      • Clinical accuracy support: Reduces diagnostic errors when used as a second opinion with physician oversight 

Challenges

      • Hallucination: AI confidently generates false, fabricated, or outdated information presented as fact 
      • Bias in training data: AI systems trained on biased data perpetuate and amplify those biases in outputs 
      • Data privacy risks: AI tools may retain, share, or train on sensitive data entered by users 
      • Lack of transparency: AI decision-making is often a "black box" - hard to audit or explain to patients 
      • Workforce & ethical concerns: Over-reliance on AI risks displacing human judgment in high-stakes decisions 

 

AI Data Classification

Our data classification system follows the guide provided by Texas Department of Information Resources.

globe icon on green to represent public data

Public

The Public label is used for information such as published reports, press releases, and information published to TTUHSC's public website. Such public-related materials require no authentication and are freely distributable by all TTUHSC personnel and are available for citizen access without requiring intervention by TTUHSC employees.

ID badge on amber representing Sensitive Data

Sensitive

The Sensitive label is used for information that may be subject to disclosure under the Texas Public Information Act, but should be vetted/verified before it is released. While these records and information are considered Public under the Texas Public Information Act, they should still be afforded a higher level of protection to ensure Confidential Data (e.g., net salary information) is not commingled. Sensitive Data may include Confidential Data that has not yet been classified as such. 

Locked safe on orange representing confidential data.

Confidential

The Confidential label is used to identify information that TTUHSC collects and maintains that is protected from disclosure either through a codified exception to the Public Information Act, Texas Government Code Ch. 552 or through the opinions or decisions of the Attorney General's Public Information office. Such information may also be subject to breach notification requirements under Texas law. 

Shield with a locked pad lock on red representing regulated data

Regulated

The Regulated label is used to identify information that TTUHSC collects and maintains that is controlled by state or federal law, and other constitutional, statutory, judicial, and legal agreements and requirements. Authorized disclosure or release of regulated data is governed by applicable statutes. 
 

View Texas Data Classification Guide

 

 

AI Tool Use Matrix

Here's a quick checklist of items you need to check off before getting started with any of these AI technologies: 

  • Complete all required AI training 
  • Read and understand your organization's AI policy 
  • Identify your specific use case and confirm that it is appropriate 
  • Consult with your supervisor before proceeding 
  • Verify that the tool you plan to use appears on your organization's approved list 
  • Understand data input, privacy, and security requirements 
  • Establish verification and quality control procedures 
AI Services Public Sensitive Confidential/Regulated
      HIPAA FERPA SSN PCI EC IRB CUI
Microsoft Copilot (TTUHSC License) 
YES
YES
YES
YES
YES
YES
YES
YES
YES
OpenAI ChatGPT Edu (Departmental purchased) 
YES
NO
NO
NO
NO
NO
NO
NO
NO
Grammarly (Departmental purchased) 
YES
NO
NO
NO
NO
NO
NO
NO
NO
Any Personally Paid AI Tools (Grammarly, ChatGPT, Claude, Copilot, Gemini) 
YES
NO
NO
NO
NO
NO
NO
NO
NO
Free Models                   
OpenAI ChatGPT (Free) 
YES
NO
NO
NO
NO
NO
NO
NO
NO
Anthropic Claude (Free) 
YES
NO
NO
NO
NO
NO
NO
NO
NO
Google Gemini (Free) 
YES
NO
NO
NO
NO
NO
NO
NO
NO
Other Free AI Tools (Apple Intelligence, Free Grammarly) 
YES
NO
NO
NO
NO
NO
NO
NO
NO

  • Published Data - Generally Low Risk
  • Controlled Data - Generally Moderate Risk
  • Confidential Data - High to Very High Risk Depending on the Type of Data and the Compliance
  • Framework it Falls Under, including:
  • HIPAA - Health Insurance Portability and Accountability Act
  • FERPA - Family Educational Rights and Privacy Act
  • SSN - Social Security Numbers
  • PCI - Payment Card Industry (i.e., Payment Services - Credit Cards)
  • EC - Export Controlled Data
  • IRB - Institutional Review Board
  • CUI - Controlled Unclassified Information 

 

Help Us Configure Co-Pilot

We're continually refining Copilot's configuration. If this message appears, please follow this link to provide feedback.

Submitted information will be reviewed by IT Security to evaluate potential Copilot policy updates in accordance with TTUHSC security, compliance, and data protection requirements.

Access Copilot Feedback Form

Screenshot of text that reads: This request couldn't be fully completed due to your organization's policies. Learn about access restrictions

Prohibited AI Models

Texas Department of Information Resources

Per Texas Department of Information Resources (Texas DIR), AI models associated with these vendors may not be used or installed on state devices. Please visit this link for the most up-to-date list of prohibited AI models from Texas DIR: https://dir.texas.gov/information-security/covered-applications-and-prohibited-technologies 

*This information is subject to change. 

  • Qwen by Alibaba  
  • WuDao by BAAI 
  • Baichuan LLM by Baichuan 
  • ERNIE by Baidu 
  • Doubao by ByteDance 
  • DeepSeek LLM by DeepSeek 
  • SparkDesk by iFlytek 
  • abab models by MiniMax 
  • Kimi by Moonshot AI 
  • Step LLM by StepFun 
  • Hunyuan by Tencent 
  • GLM / ChatGLM by Zhipu/Z.ai 

 

Explore our Latest AI Updates

Copilot icon on a sign with blurred microsoft icons in the backgroudn

Watch for our Upcoming Webinar: Introduction to Copilot Chat

We are actively in the planning to setup an interactive webinar on maximizing Copilot Chat in your daily workflows. Whether you're looking to streamline your tasks or explore new AI capabilities, this session will provide actionable insights to help you work smarter. Check back for details and ways to register.

Microsoft Copilot is Now Available

Our latest blog post is now live! Dive in to learn how Microsoft Copilot can help streamline your daily tasks, boost your productivity, and transform the way you work with AI. 

Prompt Writing for Copilot Chat

Want to get better results from your AI interactions? A new Knowledge Base article is now available, offering a step-by-step guide on how to write effective prompts for Copilot Chat to help you get the exact answers you need.

AI Training & Resources HUB

Master AI toolsets safely, efficiently, and in compliance with state guidelines.

  • Required Training | Complete the 2026 Responsible AI Use Training on your ACME dashboard by Friday, July 31, 2026 (Required per Texas House Bill 3512).
  • Quick-Start Articles | Access institutional guides to deploy Copilot Chat across Teams, Outlook, Word, Excel, and PowerPoint using your standard login credentials.
  • Advanced Learning | Access self-paced modules via Microsoft Learn, video tutorials, educator-specific frameworks, and unlimited on-demand courses through Udemy.

Check Out the Training Hub for AI

 

AI Resources/References