Quick Buzz Feed
The University of Arizona offers a comprehensive guide to various artificial intelligence tools available to its employees, detailing their diverse capabilities for everyday tasks, research, and teaching. These university-supported platforms aim to provide a secure and protected environment for data, giving faculty and staff access to a range of AI options without needing individual subscriptions.
U of A GenAI provides access to multiple large language models through one secure chat interface. Users can compare responses from different AI models, upload files, search the web and use the tool to support research, teaching and everyday work. The platform currently offers models from Anthropic Claude, Google Gemma, OpenAI GPT, Meta Llama and Amazon Nova, with model availability subject to change as new options emerge. Example uses include brainstorming, answering questions, drafting content, summarizing documents, analyzing files and charts, prototyping workflows, developing departmental chatbots, and assisting with course materials.
The AI Companion is built directly into Zoom, and employees can enable it within their U of A Zoom account settings. Due to the varying sensitivity of meeting content, users are advised to consider the meeting's nature before automatically enabling AI Companion. Example uses include catching up on meetings if arriving late, summarizing meetings and chat threads, identifying action items, drafting chat messages, and organizing cloud recordings into chapters and highlights.
Microsoft Copilot is an AI assistant seamlessly integrated with Microsoft products widely used by employees, such as Outlook, Teams, Word, PowerPoint, and Excel. Its example uses include summarizing Teams meetings, creating meeting minutes, drafting emails, developing presentations, and analyzing data.
As an Adobe Creative Campus, faculty, staff, and students at the university have access to generative AI tools within Adobe Creative Cloud, including Firefly, Photoshop, Express, and Illustrator. Firefly’s AI model generates images using Adobe Stock images, a method designed to mitigate copyright concerns. Example uses for these tools include generating and editing images and videos, removing backgrounds, and modifying specific elements within images.
Gemini is an AI assistant accessible on the web and through Google applications like Gmail, Docs, Sheets, and Drive. Its capabilities include analyzing and summarizing documents, brainstorming ideas, drafting communications, and developing content in multiple languages.
Gemini Notebook is an AI research tool that processes information from user-provided sources such as PDFs, Google Docs, and other documents. Example uses include summarizing and comparing documents, organizing information, exploring ideas, preparing for meetings, and supporting tasks like accreditation reviews, strategic planning, and policy analysis.
Scheduled for release in the fall semester, the AI Discovery Lab platform will support teaching and research by providing dedicated sandboxes for executing artificial intelligence and machine learning tasks. Example uses include analyzing and visualizing data, accessing computing resources, and supporting workshops and teaching labs.
Pilot testing for the Arizona AI Assistant platform will commence in the fall semester. This tool will enable users to create and manage dedicated knowledge databases and leverage AI models specifically tailored for these databases. Example uses include creating custom models for research and agentic tasks.
The AI Classroom (AI Verde) platform is currently undergoing testing and is expected to be released in the fall semester. It will allow users to develop state-of-the-art AI models for their classes, with example uses centered around creating AI models for specific courses.
University-supported AI tools come with enhanced privacy and security protections; however, not all university data types are compatible with every tool. Before inputting any university information into an AI tool, users should consult the AI Tools and Data Use Guide to ascertain appropriate usage based on data sensitivity. Darcy Van Patten, executive director of specialized services and digital resources at University Information Technology Services, emphasizes that a primary consideration when evaluating AI tools is their ability to prevent university data from being shared with large language models. Public information, such as published research, press releases, policies, and event details, is generally safe for AI tools. Internal data, including memos, budget drafts, and other non-restricted information, should only be used in university-supported AI tools accessed with a NetID. Restricted data, such as FERPA-protected student records, HIPAA health data, Social Security numbers, credit card information, and non-disclosed research intellectual property, must not be entered into AI chats. The Office of Responsible AI advises users to remove personally identifiable information, delete chats after use, and avoid using AI for storing or maintaining student records.
Employees have access to various resources for guidance and support in choosing and utilizing AI tools. The Office of Responsible AI provides comprehensive resources, including AI guidelines and principles, a monthly newsletter, and events such as an AI Town Hall on Sept. 2, which aims to help shape the future of AI at the U of A. The office also hosts the AI Insights Series every other Tuesday from noon to 1 p.m. on Zoom, featuring faculty, staff, and other experts discussing AI projects and applications. David Ebert, chief AI and data science officer, states their role is to help everyone understand how to use AI responsibly and ethically to enhance effectiveness and efficiency. Faculty seeking assistance with integrating AI into their courses can visit the University Center for Assessment, Teaching and Technology’s Teaching with AI resources website, explore upcoming UCATT AI events, or schedule a consultation with Adele Leon, senior instructional designer at UCATT.