AI Tools and P-Data Levels
Get to know the AI Tools available to you and the data levels that each tool can be used for. If you’re unsure, treat data as more sensitive and ask for guidance.
| Tool | Security | Best for | How to access |
| Google Gemini | P3 | Moderate | Daily drafting, brainstorming, and Workspace integration | gemini.google.com (sign in with CalNet) |
| Gemini Gems (NEW) | P3 | Moderate | Creating custom “personas” or instructions for recurring tasks | gemini.google.com (sign in with CalNet) |
| NotebookLM | P3 | Moderate | Deep research grounded in your uploaded documents | notebooklm.google.com (sign in with CalNet) |
| Berkeley AI Sandbox | P3 | Protected | Testing multi-models (GPT-4o, Claude) and images | ai-sandbox.berkeley.edu (sign in with CalNet) |
| Playlab.ai | P3 | Moderate | Role-based custom bots, forms, and voice inputs | playlab.ai (email [email protected]) |
| Miro AI | P3 | Moderate | Visual collaboration, clustering sticky notes, and diagramming | miro.com (email [email protected]) |
|
Tier |
Classification |
AI Usage Guidance |
|
P1 |
Public |
Approved for general AI use |
|
P2 |
Low / Institutional Internal |
Approved tools with institutional protections |
|
P3 |
Moderate / Sensitive |
Enterprise-grade or Haas-approved tools only |
|
P4 |
Restricted |
NO AI processing allowed |
The Human Edge Checklist
AI can provide perspective, assistance, and inform decisions – it doesn’t replace the value that humans provide. Use this checklist as a reminder for responsible use of AI, including ensuring quality, safety, and accountability.
Own Your Work, “No AI Workslop”: Never publish AI output without your human review and revision – Value others’ time like you value your own. You are responsible for the outputs you share.
Ensure Focus: You set direction and strategy, so be sure you remain clear with your goals as you work with AI. It can help draft, summarize, and even research — You drive the intent, make the decisions, and provide the final output.
Converse with AI: Don’t just ask AI for answers. Talk with AI as though it is a hyper-knowledgeable assistant, giving it context and asking it for alternative perspectives and approaches to produce high-quality output.
Verify Before You Amplify: If what you generate with AI has impact, especially if it affects people’s quality of life, policy, budgets, or the school’s reputation – Validate claims, sources, and numbers before sharing.
Help Desk
Have questions? Need to submit a ticket? Send an email to [email protected].