TulaneAI

TulaneAI gives faculty and staff access to leading generative AI models and a growing library of ready-to-use agents designed for common university needs.

You can use TulaneAI to:

• Draft and refine communications
• Summarize documents and information
• Interpret policies and other materials
• Extract text from scanned files
• Improve prompts for stronger AI results
• Explore new ways AI can support your work

TulaneAI is built on Microsoft Azure AI Foundry and hosted within Tulane's own cloud environment. Your prompts and documents are not retained by or used to train the underlying AI models, and the platform operates under Tulane's existing data security policies and safeguards.

Artificial intelligence is changing how we work, teach, learn and conduct research, while creating new opportunities to accelerate discovery and address complex problems. 

It also presents new questions and responsibilities. TulaneAI gives our community a secure, university-supported environment for exploring these tools while helping protect the privacy of our community, the university's intellectual property and the value of Tulane innovation.

Human at the Helm 

AI should support human expertise and judgment, not replace them. Every AI-generated output requires human review before it is used, shared or acted upon. 

Free consumer AI tools, including free versions of ChatGPT, Gemini and similar platforms, remain prohibited for Tulane business. TulaneAI provides a secure alternative.

For the month of September 2026, every TulaneAI user will receive a limited monthly usage quota. 

For most users, this should provide enough capacity for regular, active use. More powerful AI models consume the quota more quickly, so selecting the model or capability appropriate for your task can help you make the most of your available usage. 

September will serve as an initial learning period. Tulane will review actual usage patterns across the university and use that information to establish general quotas going forward. 

There are no surprise charges. If you reach your quota, your access will simply pause until the next monthly period. Neither you nor your department will be automatically billed. 

Tulane is also developing a process through which individuals, departments and labs can purchase additional usage when needed. Additional information will be shared as that process becomes available.

Responsible use of AI begins with understanding both what you are asking an AI system to do and what information you are providing to it.

Tulane uses governance tiers to determine the appropriate level of review for an AI use case and data classifications to determine how university information may be used.

AI governance tiers 

Tier 0: Personal productivity 

Everyday productivity uses that do not involve sensitive data. 

Approval: None required. 

Tier 1: Limited sensitive or shared use 

Generative AI involving some sensitive data or shared output within a single department. 

Approval: Make your supervisor aware. 

Tier 2: Cross-department or shared workflows 

Uses involving internal data, multiple departments or shared workflows. 

Approval: Supervisor awareness, department head approval and CISO review. 

Contact TulaneAI@tulane.edu for guidance. 

Tier 3: Higher-risk or autonomous uses 

Uses that modify data, have financial or privacy implications, or involve autonomous AI agents. 

Approval: Full review involving the CISO, Office of General Counsel, Data Security and the appropriate department.

Contact TulaneAI@tulane.edu for guidance. 

Know your data 

Tulane's Data Classification Policy defines four levels of university information. Before providing information to an AI tool, understand how the data is classified and whether the tool is approved for that use.

 Level 1: Public

Examples: Published content, marketing materials, public websites and press releases. 

AI use: Approved enterprise AI tools, including TulaneAI. 

Level 2: Internal 

Examples: Memos, process documents, meeting notes and internal reports. 

AI use: Approved enterprise AI tools only, including TulaneAI and Microsoft Copilot. 

Level 3: Confidential

Examples: Financial data, strategic plans, employee records, student personally identifiable information protected by FERPA and donor information. 

AI use: Must be submitted for review and approval. 

Contact TulaneAI@tulane.edu for guidance.

Level 4: Restricted 

Examples: Social Security numbers, HIPAA-protected health information, payment card information and research subject data. 

AI use: Treat as prohibited for use in AI tools until an approved Central IT-hosted model has been confirmed for the use. View Tulane's Data Classification Policy

Artificial intelligence has an environmental cost. Generative AI requires electricity for computation and water to support data-center cooling, and Tulane believes those impacts should be acknowledged plainly. 

At the same time, AI is rapidly reshaping research, teaching and the way we work. As a university deeply engaged in the study of climate, coastal loss and environmental justice, Tulane has a responsibility to approach both realities deliberately. 

Why a university-supported platform matters 

The alternative to a centrally governed platform is not necessarily less AI use. It can mean dozens of separate tools, duplicate subscriptions and personal accounts operating without university-wide visibility. 

Consolidating AI activity through TulaneAI gives the university a clearer understanding of how these technologies are being used. That visibility makes it possible to measure usage, identify opportunities for greater efficiency and make informed decisions about our environmental impact. 

TulaneAI also operates on hyperscale cloud infrastructure rather than infrastructure hosted locally on campus.

Measuring what we can 

Platform-level consumption will be tracked so Tulane can understand usage trends over time. The environmental impact of individual AI queries can be difficult to quantify, and published estimates vary widely. Rather than relying on figures we cannot confidently defend, we will focus on what we can measure and will be transparent where uncertainty remains.

Use AI intentionally 

Individual choices also matter. We encourage members of the Tulane community to: 

• Choose the capability the task requires. The most powerful model is not always necessary for the job. 
• Prompt with intent. Clear prompts can reduce unnecessary rounds of regeneration. 
• Reuse what works. Develop effective prompts and approaches that can be adapted rather than recreating the same work repeatedly. 

Human at the Helm is not only about accuracy and judgment. Environmental consequences are also a product of the choices we make about how we use these tools.