Artificial intelligence (AI) tools have become indispensable in various domains, enhancing efficiency, automating tasks, and solving complex problems. AI tools have the power to significantly change the way you work and how you create content.
Unlike tools that merely enhance efficiency, AI can uncover hidden patterns and provide insights, predictions, and even full drafts. However, before fully embracing this newfound transformational power and using these AI tools, learn more about how to use AI tools responsibly, including how university data must be protected.
Text-based AI tools
These tools are designed to understand and generate human-like responses to text-based, natural language prompts. They can answer your questions in a conversational way, generate both text and code, perform language translation, and create diverse forms of content and layouts.
Tool | Overview | Availability | University Data Classification Level |
Microsoft Copilot | Text-based AI chatbot; also capable of generating images (formerly Bing Chat.) | Available at no additional cost to all current students, faculty, and staff when signed in with your UT Microsoft 365 (M365) account. Visit https://microsoft365.utexas.edu/microsoft-copilot for more information. See AI Assistants below for the paid version of Copilot for Microsoft 365. | Public and Controlled |
OpenAI ChatGPT | Text-based AI chatbot | For ChatGPT-specific services, controlled or confidential data usage only applies to the dedicated UT instances of ChatGPT Enterprise covered by UT Austin contracts. The free or departmental team accounts are not covered; only published data may be used with these versions. A limited number of enterprise licenses are available; reach out to security@utexas.edu for more information. | Varies |
Google Gemini | Text-based AI chatbot; also capable of generating images (formerly Google Bard.) | There is no enterprise contract but departments can pay for individual user licenses or users can sign up for free accounts. | Public |
AI Assistants
These tools provide AI-enabled assistants in productivity and collaboration tools.
Tool | Overview | Availability | University Data Classification Level |
Copilot for Microsoft 365 | AI add-on for Microsoft 365 products. Embedded in M365 apps to create, edit, and improve your content and layouts; will work across all your M365 business data to surface the information and insights you need. | The pilot is closed to new participants. The plan is to make an announcement about general availability in the coming weeks. Licenses are $378/year/user in fiscal year 2024, prorated for mid-fiscal-year purchases. | Public, Controlled, Confidential (HIPAA and FERPA only) |
Zoom AI | AI assistant that uses meeting transcripts and chats to generate content, summaries, and ideas for prioritization. | In contract negotiations for an enterprise license. | TBD |
Grammarly | AI writing assistant that understands the context of your communication so you can generate high-quality, relevant content. Can adjust text for formality, tone, and length. | There is currently a small, no-cost pilot facilitated by the Academic Affairs Office of Academic Technology. | TBD |
Developer and Researcher Tools
API access enables developers to integrate Large Language Models (LLMs) into their own applications, products, or services.
Tool | Overview | Availability | University Data Classification Level |
Azure OpenAI | Developer access to Azure OpenAI and other cognitive services (e.g., Whisper, DALL-E). | In contract negotiations to add privacy and security policy addendums to existing enterprise agreements to allow departments to purchase individual licenses. | TBD |
Google Vertex | Developer access to Google’s AI models. | In contract negotiations to add privacy and security policy addendums to existing enterprise agreements to allow departments to purchase individual licenses. | TBD |
AWS Bedrock | Developer access to third-party large language models (e.g., Anthropic, A21 Labs, Stable Diffusion) via API in Amazon Web Services (AWS). | A Generative AI/ML security addendum was added to the enterprise agreement which now allows for use of AWS generative AI and ML web services on data up to confidential. | Public, Controlled, Confidential |
AWS Sagemaker | Managed AI/ML model access and training. | In contract negotiations to add privacy and security policy addendums to existing enterprise agreements to allow departments to purchase individual licenses. | TBD |
UT-Specific Generative AI Tool
Information Technology Services, Texas Advanced Computation Center (TACC), and Academic Affairs, along with partners across campus, currently are working on a proposal to build a UT-specific large language model (LLM) based on University data and using open source build tools, with the plan to launch a UT-specific generative AI tool in Fall 2024. Stay tuned for more information.