Mindket
BeginnerFree

Spot AI fakes: images, deepfakes and voice scams

A black banner titled "Spot AI fakes" over the line "Fake images, deepfakes and voice scams".

This free course teaches four checks for what reaches you online. Is a picture made by AI? What do you do about a deepfake video or a cloned voice? What can Content Credentials and SynthID prove, and how do AI scams work? It needs no code. Each lesson gives steps you can follow today and says where the check stops working.

What you'll be able to do

  • Check a doubtful picture with a reverse image search and a close look, and know why a missing flaw proves nothing.
  • Answer an urgent call or clip by hanging up, reaching the person yourself and using a family code word.
  • Read a file's Content Credentials and run a SynthID check, and know what a missing mark does not tell you.
  • Run five checks on a message that asks for a login, a payment or your trust, and know what to do after a click.
All courses

Who it's for

For anyone who sees pictures, videos and messages online and wants to know which to trust, and for people who help relatives with their phones. You need no code. Lesson 3 offers an optional command line tool.

Lessons

  1. 1How to tell if an image is AI-generated: five checks and their limitsHow to tell if an image is AI-generated: five checks to run today, what a reverse image search finds, and why the old tells such as hands no longer settle it.7 min
  2. 2How to spot a deepfake: video, cloned voices and the family code wordHow to spot a deepfake video or a cloned voice: why eyes and ears fail, the call-back and code word checks that still work, and what detector tests found.6 min
  3. 3Content Credentials (C2PA) and SynthID watermarks: how to check them and what they proveContent Credentials and SynthID watermarks explained: how to read them with Verify, c2patool and the SynthID Detector, and why a missing mark proves nothing.7 min
  4. 4AI scams: how criminals use AI and how to protect yourselfHow criminals use AI in scams, from polished recruiter messages to fake AI apps, and five checks that work however convincing the message looks.7 min

Frequently asked questions

Is this course about spotting AI fakes free?

Yes. Every lesson is free to read. The checkers the lessons name, such as the Verify site and the SynthID detector, are the vendors' own pages. Lesson 3 says what Verify and c2patool need, and for the SynthID detector it can only repeat that Google DeepMind calls it available to everyone.

Can an AI image detector tell me for sure?

No. Lesson 1 and lesson 2 give the research: people were wrong on almost four in ten pictures in one study, and in another none of the commercial detectors evaluated reached 90% accuracy. Treat any single result as one clue.

What is the fastest habit to learn from this course?

Leave the message and check at the source. Lesson 2 does it for a call, with a number you saved, and lesson 4 does it for an email, by opening the service yourself.

Do I need to install anything?

No. A phone and a browser cover every lesson. Lesson 3 also shows c2patool, a command line tool, for readers who want to read a file's credentials on their own computer.

What should I learn after this course?

The prompt engineering course for beginners is next. It teaches the other half, asking an AI for what you want and getting an answer you can use, and its first lesson builds one prompt in five passes.

Share this course

Learn to ask AI well

You can now check what reaches you. The next course teaches how to ask an AI for what you want, in plain language, and get a clear answer.

Start prompt engineering

All courses · All articles