Protect privacy, safety, fairness, and responsibility

This course remembers completed sections, assessment point totals, and your last course page in this browser on this device. RayAI does not receive this progress or anything you write. Selected answers and written responses are not saved. Clearing browser data, using private browsing, or changing browser or device can remove or hide saved progress.

MODULE 6 | 40 minutes | FAST TRACK CORE + GO DEEPER

Protect privacy, safety, fairness, and responsibility

Core idea: do not solve a prompt problem by creating a data problem

Before entering information, ask whether you are allowed to share it, whether the tool may retain or use it, who can access the conversation, and whether a safer substitute would work. Tool policies and organization settings vary; check the current rules that apply to you.

Minimize before you prompt

• Remove names, contact details, account numbers, IDs, secrets, and unnecessary narrative detail.

• Replace real people and organizations with roles or fictional labels when identity is not needed.

• Use the smallest excerpt that can answer the question.

• Do not paste material you do not own or have permission to use.

• If the task requires sensitive data, use an approved system and documented workflow rather than improvising.

Replacing a name alone may not protect someone: a role, location or unusual detail can still identify them. Publicly visible information may still be personal or restricted. When unsure, use a fictional example.

Prompt injection

Prompt injection occurs when content alters an AI system's behavior in an unintended way. Direct injection comes from a user's input. Indirect injection can be hidden in a webpage, email, document, image, or tool result that the AI processes.

• Treat external content as untrusted data, not as authority over the user's goal.

• Do not allow content to request secrets, broaden permissions, change recipients, or authorize an action.

• Require a preview and explicit human confirmation before consequential actions.

• Validate outputs before passing them to code, tools, databases, messages, or external systems.

Fairness and dignity

AI can reproduce biased patterns or omit relevant perspectives. For decisions about people, ask whether the criteria are legitimate, consistently applied, explainable, and reviewed by an accountable human. Do not infer sensitive traits or use AI as a hidden ranking authority.

Use respectful language and avoid stereotypes, humiliating comparisons, or examples that treat one culture, accent, disability, age group, or level of technical confidence as the default. If a response does so, reject that framing and revise the task.

See it: protect the event volunteers

UNSAFE INPUT

SAFER APPROACH

Coconut challenge

Before your next prompt, remove one piece of information the model does not need. Explain why removing it makes the task safer without reducing usefulness.

Try it: practice the skill

What to do

1. Choose the task suggested for this module, or use another simple, low-risk task.
2. Open ChatGPT or another AI assistant in a separate tab or window.
3. For this module, remove personal or sensitive details that the task does not need before entering your prompt.
4. Enter your prompt and send it. Do not include private or sensitive information.
5. Read the AI answer.
6. Return to this course and complete the short notes below.
   If you cannot use an AI assistant now, read the worked example and write what you would try later. Skip the steps that ask you to open, send or rerun a prompt. Label your notes "supplied example" or "planned test". Describe only what the example shows; if no output is supplied, write what you would look for. Keep checks you performed separate from checks you plan. Do not invent an AI result.