Tufts University data generally falls into three categories: Public, Confidential, and Restricted. These categories are described below. It’s important to note that having access to an AI tool doesn’t mean that the data you’re using is licensed for that use. If you do not know if it is safe to share data with an AI tool, DO NOT DO IT. None of the AI tools listed on this page are authorized for use with Restricted data.  

• Public Data
  Data appropriately made available to all entities in scope or to the general public. This data includes data for which unauthorized disclosure or use would not have an adverse effect on the university's reputation, finances, operations, or individuals. Examples of public data include: information displayed on public facing web sites of the university or in university publications; research published freely and publicly; employee benefit plan coverage terms published publicly, and bulk public data is considered confidential.
• Confidential Data
  All institutional data is Confidential Institutional Data unless the data has been designated as or otherwise qualifies as either Restricted Institutional Data or Public Institutional Data. Confidential Institutional Data includes data for which unauthorized disclosure or use may have a significant adverse effect on the university's reputation, finances, operations, or on individuals. Examples of confidential data include: faculty and staff personnel files, including compensation information, emergency contact information, and personnel actions; alumni and donor personal data (other than SPI); university and department financial data, and bulk public data is considered confidential.
• Restricted Data
  Data for which the unauthorized disclosure or unauthorized use may have a severe adverse effect on the university's reputation, finances, operations or individuals. This classification includes data governed by privacy or information protection requirements articulated by law, regulation, contract, binding agreement, or industry groups. Examples of restricted data include: government issued numbers used for identification, e.g. social security numbers, protected Health Information, e.g. HIPAA protected data, financial account numbers, e.g. bank account and credit card information, and student records protected by FERPA.

For more information, consult the Information and Security and Privacy Confidentiality Policy.

Different generative AI tools are each optimized for a different main purpose. These purposes are listed below. Many tools can be used for multiple purposes; the information on this page lists the main purpose.

• Conversation-Based GenAI
  AI tools primarily accessed through a chat or conversational interface that support tasks such as answering questions, brainstorming ideas, drafting content, or explaining concepts across a wide range of topics.
• Creative Media Generation
  AI tools that generate or modify visual, audio, or video content, such as images, designs, presentations, or multimedia assets, often in response to text prompts or user inputs.
• Productivity & Workflow Tools
  AI features embedded in productivity platforms that assist with organizing work, documenting processes, managing tasks, summarizing meetings, or streamlining everyday workflows.
• Teaching & Learning Tools
  AI tools designed to meet a specific educational need, including preventing AI misuse by students, giving students immediate feedback, or enabling new kinds of assessments and activities.
• Writing & Language Tools
  AI tools designed to support writing and language tasks, including drafting, editing, rewriting, grammar checking, and improving clarity, tone, or style of text.