AI Services

Find The Right Tool for your Data

For information on how each of our AI tools handles data, please click the link here.

Using AI the Right Way

Learn how CUIT recommends using AI responsibly across academics, administration, and research, click here for more information.

Use of AI tools with Sensitive Data

Sensitive Data, as defined in the University’s Data Classification Policy, may only be used with AI tools that have been specifically vetted and approved for such use by the University. At this time, the AI tool(s) approved for use with Sensitive Data are:

• ChatGPT Education, including access through the API.

For users at CUIMC, please refer to AI and Generative Technology Use at CUIMC for additional information regarding AI tools approved for use with Sensitive Data.

AI Literacy Trainings

CUIT is offering a new AI Literacy Training Program for faculty, researchers, and administrative leaders from May through August. The program features Basic and Advanced tracks, each made up of two one-hour Zoom sessions held on Tuesdays and Thursdays, with multiple enrollment options available at no cost. Participants will build practical AI skills in prompt engineering, security, and workflow integration, with office hours and cross-campus Lunch & Learn sessions available for deeper engagement. Instructional faculty may contact the Center for Teaching and Learning for training sessions or sign up at CTL Events.

To register for one of our classes, please go here.

AI Data Classification Table

ChatGPT Education

• ChatGPT Edu, available to the Columbia University community, offers enterprise-grade security and data privacy (conversations and data are not used to train OpenAI models), along with access to OpenAI's latest models including GPT-5.4 and GPT-5.4 Thinking. Features include advanced data analysis, file uploads and document summarization, web browsing, image generation and editing, custom GPTs, and Codex-powered workspace agents that can automate multi-step workflows. For more information, click here.

Claude for Education

• Claude for Education is to the Columbia University community. Claude is an AI assistant created by Anthropic that helps you think, write, code, and solve problems. You can use Claude for teaching, learning, research, and administrative work across the university. The Education offering provides enhanced security, privacy, and institutional controls. Users have access to Anthropic’s latest Claude models, along with advanced analysis tools, file handling, image capabilities, and web search. For more information, click here.
• This service is currently not available to CUIMC users as we do not have a BAA yet.

Claude API

• Claude API enables developers and marketers to integrate advanced AI directly into their applications for content generation, data analysis, and workflow automation. It is designed for high-volume, complex tasks requiring nuanced text understanding and safety.
• To request an API key for Claude, please email us at [email protected] with a project name and budget amount.
• Claude Code is only available with an API key.

OpenAI API

• The OpenAI API is available to the Columbia University community with education security and privacy. Users will have access to the latest OpenAI models in an environment capable of handling high-volume applications and large-scale data processing. For more information, click here.

Codex

• OpenAI's Codex is now available to the Columbia University community. Codex is a cloud-based AI software engineering agent that can write, debug, review, and reason about code — working autonomously on complex tasks in isolated environments so you can focus on higher-level work. Unlike traditional code completion tools, Codex understands your code base in context, accepts instructions in plain language, and can independently carry out multi-step development tasks from start to finish To learn more about Codex, click here.

CHAT

• CHAT is Columbia University's customizable AI Chat Platform, powered by LibreChat. Designed to give the Columbia Community flexible, effective access to generative AI tools. Through a streamlined interface and support for multiple advanced AI models, CHAT enables users to perform a wide range of tasks. From helping with writing, translating languages, exploring academic concepts to assisting with administrative tasks. Users can customize their experience by adjusting model settings, bookmarking chats, and using prompt templates to meet their specific needs. For more information, click here.
   

Google Gemini

• Google Gemini is now available to the Columbia University community. This service provides access to Google’s most advanced AI models in a secure, education-focused environment. Gemini supports multimodal capabilities—working with text, images, and code—so you can handle diverse academic and research tasks in one place. For more information, click here.
   

NotebookLM

• NotebookLM is now available to the Columbia University community. NotebookLM, developed by Google, is a new AI-powered research assistant designed to support your work with enhanced knowledge management capabilities. Integrated with Google Docs, NotebookLM can help you synthesize notes, generate insightful questions, and reveal connections within complex material. Ideal for academic and professional projects, this tool aims to empower faculty, researchers, and staff to manage and explore information with greater depth and efficiency, facilitating more meaningful engagement with their content. For more information, click here.
   

AI Consultation 

• To better understand your needs and ensure our services meet your requirements, we offer consultations. Please request a consultation to discuss how our AI Services can support your specific use cases and workflows. Please email [email protected] for more information. Email [email protected] for support.

 
AI Solutions and Guidance
 

Audio Transcription

• Transcription Services: Will convert audio recordings to text with high accuracy, and include features like speaker diarization.
• Accessibility: Enhances accessibility by creating searchable text documents from audio content.
• Privacy: Ensures that transcriptions are handled securely and confidentially.

Anonymization of Text 

• Data Protection: Will remove personally identifiable information (PII) from datasets to protect privacy.
• Compliance: Aims to anonymize data to meet regulatory requirements and safeguard sensitive information.
• Further reading: "LLMs are Advanced Anonymizers"

Automated Text Mining

• Data Insights: Will extract structured quantitative data from large volumes of text data by leveraging LLMs to both generate a schema and populate it. .
• Secure Processing: Ensures the security and confidentiality of text data throughout the mining process.
• Further Reading: “TnT-LLM: Text Mining at Scale with Large Language Models”

We would love to hear from you about our AI services offerings. Please fill out this survey.