Safety, Privacy, and Responsible Use of ChatGPT
Lesson 4: Your ChatGPT Safety Playbook – Put It All Together
Lesson Objectives
By the end of this lesson, students should be able to:
- Combine the course's privacy, data, and human-judgment rules into one practical playbook.
- Run a realistic task through the playbook from start to finish.
- Recognize scams and manipulation attempts that use AI-generated content.
Lesson Content
You have learned what not to share, how ChatGPT's data controls work, and when people must make the decision. A playbook puts all of it on one page so you actually use it.
A one-page ChatGPT safety playbook has four parts:
- Before you paste: remove names, account numbers, health and legal details, passwords, and confidential business information. Use placeholders.
- Settings check: know your current history, memory, and data-training settings, and use temporary chats when appropriate.
- Before you use the answer: verify facts and numbers; route medical, legal, financial, safety, and employment decisions to qualified people.
- Red flags: urgent requests for money or credentials, links you did not expect, and messages that sound official but come from odd addresses. AI makes scams more convincing, so verify through channels you already trust.
A note on manipulated content. Some web pages and documents contain hidden instructions meant to trick AI tools (sometimes called prompt injection). When ChatGPT browses or reads files for you, unusual output such as unexpected links, requests for your data, or instructions to take actions you did not ask for is a warning sign. Stop and review.
Enroll to continue this lesson.
The preview above shows the lesson objectives and opening lesson content. Enroll to view the full lesson, complete the practice work, and take the lesson quiz.