Search UT

Approved AI Tools

This page lists the artificial intelligence (AI) tools that Information Technology & Security (ITS) has reviewed and approved for use at the university. Using an approved tool does not remove your responsibility to follow the university's AI Acceptable Use Policy and the AI Standard that applies to your role. Before you use any tool listed here, confirm that the data you plan to enter is permitted for that tool.

How to read this page

Each approved tool below includes what it is, what you may and may not use it for, and the data classifications permitted. Data classifications follow the university's information classification and data protection policy:

Microsoft Copilot (within Microsoft 365)

What it is

Microsoft Copilot is the AI assistant built into Microsoft 365. It can draft and summarize documents in Word, analyze data in Excel, build slides in PowerPoint, summarize and draft email in Outlook, and answer questions in Teams and the Copilot chat experience.

Data protection commitment

When accessed with your MyUTampa account, your prompts, the responses, and the data Copilot accesses are not used to train Microsoft's or OpenAI's foundation models. Data stays within the university's Microsoft 365 service boundary and is subject to the same access controls as your other Microsoft 365 content.

Approved uses (when using MyUTampa credentials)

• Drafting, editing, and summarizing documents, email, and presentations.
• Summarizing meetings and generating action items.
• Analyzing and formatting data you are already authorized to access in Excel.
• Working with internal and, where you are the authorized user, confidential information — because it stays inside the university tenant.

Not approved

• Do not enter confidential or regulated data (SSNs, PCI, HIPAA-covered health data, contract-restricted research).
• Do not paste data you are not authorized to access; Copilot respects existing permissions but you remain responsible for what you enter.
• Do not sign in to Copilot or the consumer Microsoft Copilot app with a personal account for university work.

Permitted data classifications

Zoom AI Companion

What it is

Zoom AI Companion is the AI assistant built into Zoom. It can generate meeting summaries, capture next steps, answer questions about a meeting in progress, and help compose chat messages. It is available within the university's licensed Zoom environment.

Data protection commitment

Zoom does not use customer content — audio, video, chat, screen sharing, or attachments — to train Zoom's or third-party AI models. Note that even without model training, meeting content is still processed and generated summaries may be stored.

Approved uses

• Generating summaries and action items for internal meetings.
• Catching up when you join a meeting late ("ask AI Companion").
• Drafting routine Zoom chat and follow-up messages.

Not approved / use caution

• Do not enable AI Companion in meetings where restricted or regulated data is discussed (e.g., individual health information, disciplinary matters involving protected records). Note, you can pause the AI companion for the duration of discussions that include this classification of data.
• Always notify participants that AI Companion is on; some meetings (legal, HR, sensitive advising) should have it turned off.
• Review AI-generated summaries for accuracy before sharing — they may contain errors.

Permitted data classifications

Claude for Teams

What it is

Claude is a general-purpose AI assistant from Anthropic, licensed for university use through Claude for Teams. It can draft and revise writing, summarize and answer questions about documents you provide, analyze data, help with spreadsheets and code, and work through research and planning tasks in a chat conversation. It is available in the browser, in desktop and mobile apps, and within Microsoft 365 applications.

Data protection commitment

Under the university's agreement, Anthropic does not use university prompts, uploaded files, or Claude's responses to train its AI models. Conversations and files stay within the university's Claude for Teams workspace and are subject to the same access controls, records retention, and public-records obligations as other university content.

Approved uses

• Drafting, editing, and summarizing documents, email, and presentations.
• Summarizing, comparing, and answering questions about documents you are already authorized to access.
• Brainstorming, outlining, and planning for projects, courses, and communications.
• Analyzing data and writing or troubleshooting formulas, queries, and code.
• Working with internal and, where you are the authorized user, sensitive information — because it stays inside the university’s workspace.

Not approved / use caution

• Do not enter confidential or regulated data (SSNs, PCI, HIPAA-covered health data, contract-restricted research).
• Do not sign in with a personal Claude account, use the consumer free tier, or connect a personal API key for university work.
• Do not treat Claude's output as authoritative. Verify facts, figures, citations, and quotations before you rely on them or share them.
• Do not use Claude to reach or document final decisions about individuals (admissions, hiring, discipline, grading) without human review.

Permitted data classifications

ChatGPT Business

What it is

ChatGPT is a general-purpose AI assistant from OpenAI, licensed for university use through ChatGPT Business (formerly ChatGPT Team). It can draft and revise writing, summarize and answer questions about documents you provide, analyze data and files, help with spreadsheets and code, and work through research and planning tasks in a chat conversation. It is available in the browser and in desktop and mobile apps.

Data protection commitment

Under the university's agreement, OpenAI does not use university prompts, uploaded files, or ChatGPT's responses to train its AI models. Conversations and files stay within the university's ChatGPT Business workspace and are subject to the same access controls, records retention, and public-records obligations as other university content. Note that submitting feedback on a response (for example, a thumbs up or down) may share that conversation with OpenAI.

Approved uses

• Drafting, editing, and summarizing documents, email, and presentations.
• Summarizing, comparing, and answering questions about documents you are already authorized to access.
• Brainstorming, outlining, and planning for projects, courses, and communications.
• Analyzing data and writing or troubleshooting formulas, queries, and code.
• Working with internal and, where you are the authorized user, sensitive information — because it stays inside the university’s workspace.

Not approved / use caution

• Do not enter confidential or regulated data (SSNs, PCI, HIPAA-covered health data, contract-restricted research).
• Do not sign in with a personal ChatGPT account (Free, Plus, or Pro) or connect a personal API key for university work.
• Do not treat ChatGPT's output as authoritative. Verify facts, figures, citations, and quotations before you rely on them or share them.
• Do not use ChatGPT to reach or document final decisions about individuals (admissions, hiring, discipline, grading) without human review.

Permitted data classifications

Turnitin AI Detection

What it is

Turnitin AI Detection is a feature of the university's licensed Turnitin service. When a student submission is processed, Turnitin reports an AI writing indicator — an estimate of how much of the text may have been generated by an AI tool — alongside the standard similarity report. It only analyzes work that is submitted to Turnitin; it does not generate content or answer prompts.

Data protection commitment

Submissions are processed inside the university's licensed Turnitin environment under the university's agreement, and Turnitin does not use university submissions to train generative AI models. Student submissions, similarity reports, and AI indicators are part of the education record under FERPA and must be handled accordingly.

Approved uses

• Reviewing the AI writing indicator as one signal among several when you have a concern about a submission.
• Discussing a flagged submission with the student before drawing any conclusion.
• Documenting a suspected academic integrity concern through the university's academic integrity process.
• Supporting conversations with students about appropriate and disclosed AI use in your course.

Not approved / use caution

• Do not treat the AI indicator as proof of misconduct. The percentage is an estimate and can produce false positives, particularly for multilingual writers and for heavily edited, formulaic, or template-based writing.
• Do not base a grade penalty or an integrity finding on the indicator alone; human review and the academic integrity process are required.
• Do not share a student's report or AI indicator with anyone who does not have a legitimate educational interest in it.

Permitted data classifications

Tango AI

What it is

Tango is an AI-powered documentation tool that turns a task you perform on screen into a step-by-step how-to guide. As you click through a process, Tango automatically captures each step and screenshot and assembles a polished, shareable guide — useful for training materials, standard operating procedures, and software walkthroughs.

Data protection commitment

Tango’s contract is not managed by ITS. ITS has not verified — and has no way to confirm — whether Tango uses university content to train its AI models. The functional owner of the application is responsible for ensuring the agreement addresses AI model training and data handling. Remember that Tango captures your screen: anything visible when you record a step — names, IDs, email content — becomes part of the guide, so what Tango stores is determined by what you show it.

Approved uses

• Creating step-by-step guides and standard operating procedures for university software and business processes.
• Building training and onboarding materials for departmental workflows.
• Documenting IT and administrative procedures for knowledge bases and service desk use.

Not approved / use caution

• Do not capture screens that display confidential or regulated data (SSNs, credit card numbers, health information); use test data, or blur and redact before sharing.
• Do not record workflows in systems containing student or employee records unless the visible data is limited to what the guide’s audience is authorized to see.
• Review every captured screenshot before publishing a guide — Tango captures exactly what is on screen.

Permitted data classifications

Otter.ai

What it is

Otter is an AI meeting assistant that records, transcribes, and summarizes meetings. It can join Zoom and Teams meetings or capture in-person conversations, generate a live transcript, produce summaries and action items, and let you search across your past meeting transcripts.

Data protection commitment

Under the university’s agreement, Otter does not use university conversations to train its AI models. Note that recordings and transcripts are stored in Otter’s cloud environment — treat a transcript with the same care as the meeting it came from.

Approved uses

• Transcribing and summarizing internal meetings where participants have been notified.
• Generating action items and searching your own past meeting transcripts.
• Producing transcripts to improve accessibility for meeting participants.

Not approved / use caution

• Do not record meetings where restricted or regulated data is discussed (e.g., individual health information, legal matters, HR or disciplinary discussions).
• Always notify participants that Otter is recording and transcribing — Florida is an all-party consent state, so every participant must consent before recording.
• Review AI-generated summaries for accuracy before sharing — they may contain errors.

Permitted data classifications

Grammarly

What it is

Grammarly is an AI writing assistant that checks spelling, grammar, tone, and clarity as you write, and can rewrite, shorten, or draft text on request. It works in the browser, in Microsoft Office, and in email, offering suggestions wherever you type.

Data protection commitment

Grammarly’s contract is not managed by ITS. ITS has not verified — and has no way to confirm — whether Grammarly uses university text to train its AI models or how it otherwise handles university data. The functional owner of the application is responsible for ensuring the agreement addresses AI model training and data handling. Keep in mind that Grammarly processes the text of whatever document or message it is active in — where you enable it determines what it sees.

Approved uses

• Proofreading and improving the clarity, tone, and grammar of documents, email, and web content.
• Rewriting, shortening, or adjusting the tone of drafts you provide.
• Supporting consistent, professional writing across departments.

Not approved / use caution

• Do not enable Grammarly in documents or systems containing confidential or regulated data (SSNs, PCI, HIPAA-covered health data) — it processes the text it can see.
• Do not use a personal or free Grammarly account for university content.
• Do not rely on Grammarly’s suggestions for high-stakes or legally sensitive wording without human review.

Permitted data classifications

Enflux

What it is

Enflux is a data analytics and academic decision-support platform for health-science programs. It brings exam and assessment results together with curriculum data to track student performance, predict board-exam (PANCE) readiness, classify exam items against competencies, and support accreditation reporting.

Data protection commitment

Enflux’s contract is not managed by ITS; the Physician Assistant Medicine program is the functional owner. ITS has not verified — and has no way to confirm — whether Enflux uses university or student data to train its AI models; the program is responsible for ensuring its agreement addresses AI model training and data handling. Student academic records processed in Enflux remain protected by FERPA and are limited to program faculty and staff with a legitimate educational interest.

Approved uses

• Analyzing exam and assessment performance for students in the Physician Assistant Medicine program.
• Monitoring cohort progress and predicting board-exam readiness.
• Classifying exam items against competencies and supporting accreditation reporting.

Not approved / use caution

• Do not upload data beyond what the platform requires — student academic records only; no health, financial, or disciplinary information.
• Do not share Enflux dashboards or reports with anyone who lacks a legitimate educational interest in the student data they contain.
• Do not treat predictive scores as the sole basis for a decision about an individual student — use them alongside faculty judgment.

Permitted data classifications

Google NotebookLM

What it is

NotebookLM is Google’s AI research and note-taking assistant. You upload your own sources — documents, slides, PDFs, web pages — and NotebookLM answers questions, builds summaries, study guides, and audio overviews grounded only in those sources, with citations back to them.

Data protection commitment

The Google agreement under which NotebookLM is used is not managed by ITS. ITS has not verified — and has no way to confirm — whether Google uses NotebookLM content to train its AI models. The functional owner of the application is responsible for ensuring the agreement addresses AI model training and data handling. Content you upload is stored with Google, so the data classification rules apply to every source you add.

Approved uses

• Summarizing and asking questions about documents you are already authorized to access.
• Creating study guides, briefing documents, and audio overviews from your own uploaded sources.
• Organizing research and course materials into searchable, citable notebooks.

Not approved / use caution

• Do not upload confidential or regulated data (SSNs, PCI, HIPAA-covered health data, contract-restricted research).
• Do not sign in with a personal Google account for university work.
• NotebookLM answers only from the sources you give it — verify important facts against the originals, and remember its citations point to your uploads, not the wider web.

Permitted data classifications

Mentimeter

What it is

Mentimeter is an interactive presentation and audience-engagement tool. Presenters build slides with live polls, quizzes, word clouds, and Q&A, and participants respond from their phones or laptops, by name or anonymously. Its AI features can draft presentations and questions from a prompt and group or summarize open-ended audience responses.

Data protection commitment

Mentimeter’s contract is not managed by ITS. ITS has not verified — and has no way to confirm — whether Mentimeter uses university content or participant responses to train its AI models. The functional owner of the application is responsible for ensuring the agreement addresses AI model training and data handling. Keep in mind that Mentimeter stores your presentations and every audience response — whatever participants type becomes part of results that can be viewed, shared, and exported.

Approved uses

• Running live polls, quizzes, word clouds, and Q&A in classes, meetings, trainings, and events.
• Gathering audience feedback and checking understanding during a session.
• Using the AI features to draft presentation content and questions, or to group and summarize open-ended responses.

Not approved / use caution

• Do not include confidential or regulated data (SSNs, PCI, HIPAA-covered health data) in slides or AI prompts, and do not ask participants to submit it.
• Do not use a personal or free Mentimeter account for university content.
• Review AI-generated content and response summaries for accuracy before presenting or sharing — they may contain errors.

Permitted data classifications

HeyGen Creator

What it is

HeyGen is an AI video-generation platform. From a written script, it produces presenter-style videos featuring stock or custom AI avatars and synthetic voices, and it can translate and dub existing videos into other languages with lip-synced audio. The Creator plan is a single-user plan for producing finished videos up to 30 minutes long at 1080p, and it includes one voice clone.

Data protection commitment

HeyGen’s subscription is not managed by ITS. ITS has not verified — and has no way to confirm — whether HeyGen uses university content to train its AI models; HeyGen’s published materials describe its practices inconsistently, and opting out of model training may require a direct request to HeyGen. The functional owner of the application is responsible for ensuring the subscription terms address AI model training and data handling. Keep in mind that HeyGen stores your scripts, uploaded media, and any avatar or voice clone created from a person’s image or voice — a likeness is identifying personal information and should be handled with the same care.

Approved uses

• Producing training, onboarding, and instructional videos from scripts you write.
• Creating short explainer, announcement, and outreach videos for public-facing university content.
• Translating and dubbing existing university videos to reach multilingual audiences and improve accessibility.

Not approved / use caution

• Do not create an avatar or voice clone of anyone — including yourself, colleagues, or students — without that person’s documented consent.
• Do not include confidential or regulated data (SSNs, PCI, HIPAA-covered health data) or student or employee records in scripts or uploaded media.
• Do not use HeyGen to impersonate a real person or to produce content that could be mistaken for a genuine recording; clearly disclose that the video is AI-generated.
• Review every generated video, including translations, for accuracy before publishing — AI voices and translations may contain errors.

Permitted data classifications

Requesting a new AI tool

Because AI tools often carry contractual terms and data-handling implications, all new AI tools must be reviewed by ITS before purchase or use. To request a new AI tool, review the university's Software and Technology Equipment Approval Policy and the university’s annual budget process.

Frequently asked questions

Can I use ChatGPT, Gemini, or another public AI tool for UTampa work?

Not for institutional data unless it appears on this approved list and you’ve logged into it with your MyUTampa credentials. Public consumer AI tools often use your input to train their models and may not meet the university's contractual and privacy requirements. Submit a request to ITS if you have a business need.

Is my data safe in the approved AI tools?

For the tools managed by ITS (Microsoft Copilot, Zoom AI Companion, Claude for Teams, ChatGPT Business, Turnitin AI Detection, and Otter.ai), the university’s agreement prohibits the vendor from training its AI models on our data. For tools purchased and managed by individual departments (Tango AI, Grammarly, Enflux, Google NotebookLM, Mentimeter, and HeyGen Creator), ITS does not manage the contract and cannot confirm the vendor’s data practices — the functional owner of the application is responsible for ensuring that protection. In every case, this is not the same as "anything goes" — you must still follow the data classification rules above and not submit regulated data.

Who do I contact with questions?

Contact the ITS by emailing UTCIO@ut.edu.

Related policies and standards

• AI Acceptable Use Policy
• Software and Technology Equipment Approval Policy
• Information Classification and Protection Policy