Responsible AI

How this workspace uses AI, and what we ask of you in return.

Human in the loop
AI output is a first draft, never a final decision. Review and edit anything before you send it, publish it, or act on it.
Accuracy & verification
Language models can state wrong facts confidently. Check names, numbers, dates, quotes and legal or medical claims against a trusted source.
Privacy by default
There are no accounts and no user profiles. Your prompts, history and saved outputs stay in this browser's local storage on your device.
Fairness & bias
Models learn from human text and can reflect stereotypes. Watch for skewed tone or assumptions in anything about people, hiring or performance.
Appropriate use
Use it for drafting, summarising, planning and brainstorming. Don't use it to impersonate people, generate harmful content, or make decisions about someone without review.
Sensitive information
What not to paste into a prompt.
  • Passwords, API keys or banking details
  • Identity numbers, medical records or other personal data about others
  • Confidential client, employer or exam material you are not allowed to share