Telling truth from AI fakes

About this course

This course teaches you to recognize synthetic content — generated images, video, text, and voices — and to check what's actually trustworthy online. It addresses a problem that now touches everyone: AI-generated fakes have grown so common and convincing that gut instinct alone no longer protects you.

The course moves from the psychology of why we believe fakes in the first place to hands-on practice: spotting signs of generated images and deepfakes, reverse image search for origin, checking text and sources against a source-pyramid method, and verifying numbers and quotes. A separate block covers voice cloning and targeted phishing, and the final project has you build a personal two-minute verification kit and test it on a blind set of five images.

The course suits anyone who reads news, watches video, or interacts online — no technical background is needed. What helps is honest willingness to examine cases where you personally believed something false, and to avoid treating AI detectors as the only verification tool.

What you'll learn

• Recognize the signs of a generated image or deepfake
• Run a reverse image search and verify a material's origin
• Evaluate sources using a source-pyramid method to find the primary one
• Verify numbers, studies, and quotes for authenticity
• Recognize voice cloning and targeted phishing
• Tell inflated reviews and comments apart from genuine ones
• Build a personal two-minute verification kit for any material
• Know what to do if you've already believed or spread a fake

What you'll end up with

• A personal two-minute verification kit ready to use right away
• A breakdown of at least three real cases where you believed or doubted something
• Results from a blind image test, checked against the answers and your own mistakes
• The habit of checking a source before sharing anything

This course isn't for someone looking for an automated tool to flag AI content — it teaches manual verification instead, and explicitly covers why AI detectors themselves can't be trusted.