Udemy
Turn what you know into an opportunity and reach millions around the world.
Learn More
Your cart is empty.
Keep shopping
AI Safety at Work: Privacy, Scams and Deepfakes
New
11 students

AI Safety at Work: Privacy, Scams and Deepfakes

Use ChatGPT, Claude, Gemini and Copilot safely: what not to paste, the right settings, and how to beat AI scams
Created byMuhammad Shahbaz Khan
Last updated 10/2026
English

What you'll learn

  • Decide what is safe to paste into an AI chat and use redact-and-replace to protect names, numbers and confidential details.
  • Find and set the training, memory and history controls in ChatGPT, Claude, Gemini and Copilot, and know when a work account changes the rules.
  • Use temporary or incognito chats and connected-app permissions sensibly, knowing what they do and don't protect.
  • Recognise AI voice clone, deepfake video and AI-written phishing scams, and stop them with a code word and call-back habit.
  • Check AI answers for confident mistakes with a quick source check before they reach a customer or affect money.
  • Respond calmly if something sensitive is pasted, and roll out a one-page AI use policy for your team.

Explore related topics

Course content

6 sections • 22 lectures • 42m total length

  • AI Safety at Work: Start Here1:58

    By the end of this lecture you will know who this course is for and how to use the downloads. The Prompt Library and the one-page team AI policy template are attached to this lecture, so you can download them from the Resources section below.

  • What Happens to What You Type2:07

    By the end of this lecture you will understand what happens to your words after you press send: chat history, model training, human review and memory.

Requirements

  • You use, or plan to use, at least one AI chat tool such as ChatGPT, Claude, Gemini or Copilot.
  • No technical or security knowledge needed.
  • Optional: access to your own AI account settings so you can follow along.

Description

This course contains the use of artificial intelligence.

You use AI chat tools at work. Maybe you draft emails in ChatGPT, summarise reports in Copilot, or brainstorm in Claude or Gemini. But do you know what happens to what you paste in? Do you know where the training and memory settings are? And would you spot a cloned voice on the phone asking for an urgent payment?

This short, practical course is for everyday AI users, small business owners and team managers. No technical background needed.

You'll learn what should never go into an AI chat and a simple redact-and-replace habit that keeps the help without the risk. You'll get a settings tour of ChatGPT, Claude, Gemini and Copilot, with each setting described by name and checked against the vendors' own help pages. You'll see what temporary and incognito chats do and don't hide, and how to check which apps you've connected.

Then we turn to scams. You'll see how criminals use AI for voice clones, deepfake video calls and polished phishing emails, drawing on warnings from the FBI, the FTC and the UK's NCSC. You'll set up a code word and call-back habit for your family and your team, and learn where to report a scam in the UK and the US.

Finally, you'll learn to catch confidently wrong AI answers with a five-minute source check, what to do in the first hour if you paste something sensitive, and how to roll out a one-page AI use policy for your team.

Almost every lesson comes with a copy-paste prompt or fill-in template, including the full team policy.

This course is general guidance, not legal advice. AI tools change their menus often, so settings are described by name.

AI disclosure: This course was created with the help of AI. The research, structure and examples were reviewed by the instructor. The narration uses an AI-generated voice.

Who this course is for:

  • Office workers and freelancers who use AI chat tools at work and want to stay safe without becoming security experts.
  • Small business owners who want simple rules for AI, payments and scam calls.
  • Team managers looking for one short course and a ready-made AI policy for their whole team.

Report abuse