Adventure 03: Real or Fake?

You see a photo of a celebrity doing something wild. You see a “news” headline that makes your blood boil. You get a voice note that sounds exactly like someone you know. Welcome to the age of amazing fakes — where computers can cook up pictures, voices, and articles that feel real. Your mission: become a digital detective with a magnifying glass made of questions, not fear. Let’s go.

What You'll Learn

• Why AI-made images, videos, and text can look real even when they’re not.
• What deepfakes are — in plain language — and why they matter for friends, schools, and the wider world.
• Detective tricks for photos, headlines, and videos (no fancy software needed).
• How fact-checking works like a sport: sources, cross-checks, and calm thinking.
• How to protect yourself and others by slowing down before you share.

Adventure 1: The Hall of Mirrors — What AI Can Fake

Imagine a fun-house mirror that doesn’t just stretch your face — it invents a whole person who never existed, or puts a real person’s face on a fake body, or writes an article that sounds official. AI tools can:

• Images: “People” who never lived, fake event photos, impossible animals that look real.
• Video: Faces and voices matched to words someone never said (deepfakes).
• Audio: Cloned voices for pranks — or scams.
• Text: Articles, reviews, or comments that feel human — including made-up “facts.”

This isn’t about making you scared of every pixel. It’s about knowing: seeing isn’t always believing anymore.

Think about school gossip: one twisted sentence can travel the whole lunchroom. Online, a fake screenshot can do the same thing — faster. Your detective skills protect real people.

Visual: Sketch a detective badge labeled “I ask before I amplify.” Color it however you want — detectives can love glitter too.

Adventure 2: Photo Clues — Hands, Text, and “Almost Right”

AI photos are getting better, but glitches still show up — especially if you know where to peek.

Important: As AI improves, one weird hand won’t always mean fake — use several clues plus source checks.

Cool Fact!
Real sports photos sometimes look “weird” because of motion blur — that’s not the same as AI slime-fingers. Context matters!

Comparison chart — “Suspicious vs Normal”:

Adventure 3: Deepfakes & Video — When Faces Lie

A deepfake video often uses AI to map one person’s expressions onto another or to synthesize speech. Uses range from movie effects to harmful hoaxes.

Video detective habits:

• Listen + watch: Do emotions match the words? Sometimes the face is “almost” right but off.
• Lip sync: Weird timing between mouth and sounds.
• Edges: Flicker around hairline or jaw (not always visible — tech improves fast).
• Source: Did official channels post this, or a random account with no history?

Table — “Safe response if something feels off”:

Adventure 4: Text Tricks — Headlines, “Experts,” and Smooth Lies

AI text can sound smooth and confident — even when wrong. Fake news often pushes your feelings because feelings make you share before thinking.

Headline checks:

• Does the article match the headline?
• Is there a real author with other work?
• Do other trusted sources report this?
• Is the site new with all viral stories?

AI text clues (not proof alone):

• Everything feels evenly polished — no rough personal details.
• Fake citations — study names or books that don’t exist (double-check with adults).
• Vague specifics (“scientists say”) without who / where / when.

Adventure 5: The 3-Check Rule — Your Everyday Shield

Before you believe or share something huge:

1. Check the source — Who published it? What’s their track record?
2. Check other sources — Who else confirms it?
3. Check your feelings — Am I being rushed or manipulated?

Add a bonus 4th check with AI content: Check with an adult if it involves money, people’s reputations, or scary threats.

Adventure 6: Playground Rules — Rumors, Screenshots, and Kindness

A screenshot can be cropped, edited, or old. A group chat can turn into a rumor rocket. Being a detective isn’t only about AI — it’s about not hurting real humans with unproven stories.

Visual: A STOP sign made of question marks — caption: “Pause is not weakness.”

Adventure 7: Trust Toolkit — Who Counts as a “Good Source”?

Not all websites wear honest T-shirts. Libraries, schools, and scientists spend years learning how to check facts — you can borrow their habits.

Remember: Even good sources update when they learn more — science grows. Changing your mind with new evidence is smart, not weak.

Cool Fact!
Some museums and zoos have fact pages written for kids — great training ground for your source muscles.

Visual: A backpack with pockets labeled library, teacher, official site — fake news tries to rip the zipper; you zip it back.

Adventure 8: Cartoons, Games, and Group Chats — Practice in Safe Worlds

You can train your detective muscles without doom-scrolling:

• Cartoon plot twist: Was the “evil twin” really evil — or was someone framed? Notice how evidence works in stories — then use the same habit online.
• Games: If a quest rumor spreads in chat (“secret level at 3 a.m.!”), that’s practice for asking: Who tested this?
• Sports trade rumors: Even pros get fooled by fake tweets. Fans who wait for official team posts stay calmer.

Tell Someone (Share the Detective Kit)

Pick one habit from this module and teach it live:

Step 1 — Demonstrate the 2-minute pause with a made-up headline (not a real rumor).
Step 2 — Show your fake-o-meter drawing and move the dial while naming clues.
Step 3 — Ask: “When was a time you almost shared too fast?” Listen without judging.
Step 4 — Promise each other: “We’ll text ‘source?’ before we pile on.”

Quiz Time! (Truth Detective Exam)

1. Name two things AI can fake besides still photos.
Sample answer: Video/voice (deepfakes), text/articles, audio clips.

2. Why shouldn’t you panic-share when a headline makes you furious in one second?
Answer: Strong feelings are a trick that speeds up sharing before you verify — pause helps you think.

3. What’s one photo clue that sometimes (not always) signals AI weirdness?
Sample answer: Strange hands/fingers, gibberish text in the image, impossible shadows.

4. What is a deepfake in kid words?
Answer: Video or audio where a face or voice is changed or made up so it looks/sounds like someone said or did something they might not have.

5. What are the three checks in the 3-check rule?
Answer: Source, other sources, your feelings (plus adult for big/scary/money stuff).

6. True or false: If a screenshot exists, the story is automatically true.
Answer: False — screenshots can be cropped, edited, or old.

7. What is lateral reading?
Answer: Opening new tabs to learn about who made the site — not only reading down the page.

8. What should you do if a message asks for a gift card “right now”?
Answer: Stop, tell a trusted adult, and verify through a known contact method — don’t send cards because of one message.

Activities

1. Detective journal: For one week, note one “maybe fishy” thing per day (with adult OK). Write: clue / verdict / what I’d do next.
2. Poster: “Don’t Feed the Hoax”: Draw five emoji-style reactions — circle only the ones that mean pause and verify (magnifying glass, calm face).
3. Role-play newsroom: You’re the editor. Three “stories” come in — only one gets printed. Debate why with family.
4. AI art label: Create or pick one AI image (with permission). Add a big honest label: “Made with AI.” Discuss why labels help trust.
5. Kindness scenario: Write a short script where someone shares a fake rumor about a classmate. Your character stops the chain without being mean.
6. Timeline trick: Draw a real news timeline vs a fake one (jumbled dates). Label how fakes shuffle time to confuse you.
7. Interview: Ask a librarian or teacher: “What’s one tool you use to check a claim?” Draw their answer as a tool in a toolbox.
8. Ad break: Watch one TV or online ad with an adult. List three tricks it uses (happy music, urgency, famous face). Same tricks show up in sketchy posts.
9. Sound-alike test: Record yourself saying one true sentence and one silly lie (family joke). Can a listener tell? Discuss why voice clones are extra tricky now.
10. Good news hunt: Find one boring true story with two sources — celebrate boring truth like a trophy.

Your Challenge (Mini-Project)

Build a “Digital Detective Kit” (shoebox or folder).

Include:

• Your 3-check card (decorate it!).
• A one-page comic (4–6 panels): Fake Alert! Show clues you learned.
• A “Real vs Fake” mini-quiz with 5 questions for friends or family (answers on the back).
• A pledge line: “I slow down before I spread something that could hurt someone’s reputation.”

Optional with permission: one screenshot of a practice fact-check — blur all personal info.

What You Learned

• AI can create convincing fakes in images, audio, video, and text — fast.
• Glitches and sources are clues, but habits (pause, verify) matter most.
• Deepfakes can be used for fun or harm — consent and context decide a lot.
• Headlines and emotions are tools manipulators use — you can notice the pressure.
• Fact-checking is a skill you can practice like music or sports.
• You protect others when you don’t amplify unverified rumors.
• Kindness + curiosity beats panic-sharing every time.

Show Off!

• Run your mini-quiz at dinner — crown someone Chief Detective if they nail the reasoning, not just the guess.
• Present your comic to a class or club — focus on helpful habits, not scary vibes.
• Share your kit (show-and-tell style) and explain one clue you’ll never forget.
• Teach the 2-minute pause to a friend — the internet needs more thoughtful speed bumps.
• Post-it promise: Leave a sticky on your mirror: “Sources before shares.”
• Library field trip show-back: After visiting (or a virtual tour), teach one source trick you learned to two friends.
• Family challenge: Weekend contest — who finds the weirdest true fact with two sources? Winner picks dessert.

You’re not a passive scroller. You’re someone who looks closer — and that makes the whole playground (online and off) a little safer.

Imagine if your group chat had a rule: no forwarding screenshots without the original link (when safe). What would change? Draw your team’s “detective mascot” with three tools in its belt (pause, ask, verify).