StanfordUniversity IT Responsible AI for everyone
AI Catalyst

7 ways forward

Start with your goal.

Each path is short enough to finish, specific enough to be useful, and grounded in Stanford's policies and supported tools.

Path 01Developers and technical staffintermediate

Build with the API Gateway

You hold an AI API Gateway key charged to your PTA, have made a successful chat completion call to aiapi-prod.stanford.edu, and know which data classifications may pass through it.

5 steps About 1.5 hours

  1. AI API Gateway | University IT
  2. AI API Gateway FAQs | University IT
  3. KB00019978: Stanford AI API Gateway Access and Usage

and 2 more steps

Start this path

Path 02Faculty, instructors and teaching teamsbeginner

AI for Teaching & Coursework

You can state Stanford's expectations for AI use in coursework, have a Claude for Education account, and have tried a guided-learning mode on one topic from your own course before deciding whether students should.

5 steps About 2 hours

  1. Responsible AI at Stanford
  2. Ethics in the Age of Generative AI
  3. AI Fluency for Educators

and 2 more steps

Start this path

Path 03Staff and faculty new to AIbeginner

New to AI at Stanford

You have a Stanford ChatGPT Edu account, can say which data you may put into it, and have drafted and revised one real document in it with a prompt you wrote yourself.

6 steps About 1 hour

  1. GenAI FAQs | University IT
  2. Responsible AI at Stanford
  3. OpenAI ChatGPT Edu at Stanford

and 3 more steps

Start this path

Path 04Staff and researchers handling sensitive dataall levels

High-Risk Data & AI

You can name which Stanford AI services are approved for High Risk data and which for PHI, place your own use case on Stanford's AI risk tiers, and choose the right tool before you paste anything.

6 steps

  1. Responsible AI at Stanford
  2. AI Playground | University IT
  3. Microsoft Copilot Chat | University IT

and 3 more steps

Start this path

Path 05Staffbeginner

Get productive with AI at Stanford

You have one recurring piece of your own work, such as a weekly report, set up as a Project in ChatGPT Edu with your template and standing instructions, and you know which data may go into it.

6 steps About 4 hours

  1. OpenAI ChatGPT Edu at Stanford
  2. ChatGPT Edu Tutorial
  3. Responsible AI at Stanford

and 3 more steps

Start this path

Path 06Researchers and research staffintermediate

AI with research data

You can match a dataset's risk classification to an approved Stanford AI service, have requested AI API Gateway access if your data needs it, and have produced one cited literature scan you verified source by source.

6 steps About 1.5 hours

  1. Responsible AI at Stanford
  2. Know Your Risks: Agentic AI Risk Classifications | University IT
  3. AI API Gateway | University IT

and 3 more steps

Start this path

Path 07Developers and power usersadvanced

From prompts to agents

You have one agent running against the AI API Gateway with least-privilege access and a confirmation step before any external action, built with Claude Code, Codex or a no-code agent designer.

6 steps About 4 hours

  1. Responsible Agentic AI
  2. Using AI Code Assistants With AI API Gateway | University IT
  3. Agents and Workflows

and 3 more steps

Start this path

Not sure where to begin?

Tell the assistant what you want to accomplish.