Deepfakes and misinformation in AI for Everyone - Mini Project: Build & Apply

Understanding Deepfakes and Misinformation

📖 Scenario: You are learning about how deepfake videos and images can spread false information online. This project will help you understand what deepfakes are, how misinformation spreads, and how to recognize and respond to them.

🎯 Goal: Build a simple guide that explains deepfakes and misinformation step-by-step. You will create a list of examples, set a threshold for trustworthiness, identify deepfake examples, and add advice on how to stay safe from misinformation.

📋 What You'll Learn

Create a list of example deepfake scenarios with exact text entries

Add a trust threshold variable with a specific value

Use a loop to filter deepfake examples based on trustworthiness

Add a final advice statement about verifying information

💡 Why This Matters

🌍 Real World

Understanding deepfakes helps people recognize false videos and images online, reducing the spread of misinformation.

💼 Career

This knowledge is important for roles in media, education, cybersecurity, and social media management to promote truthful information.

1

Create a list of deepfake examples

Create a list called examples with these exact strings: 'Fake celebrity video', 'Altered political speech', 'Manipulated news image', 'False social media post'.

AI for Everyone

# Create the list called examples with the given entries
# Your code here

Hint

Use square brackets [] to create a list and include the exact strings separated by commas.

2

Set a trustworthiness threshold

Create a variable called trust_threshold and set it to the number 3 to represent the minimum trust level for information.

AI for Everyone

examples = ['Fake celebrity video', 'Altered political speech', 'Manipulated news image', 'False social media post']
# Set trust_threshold to 3
# Your code here

Hint

Assign the number 3 to the variable trust_threshold.

3

Identify deepfake examples below the trust threshold

Create a list called deepfake_examples that includes only the first three items from examples using a for loop and an if condition comparing the index to trust_threshold.

AI for Everyone

examples = ['Fake celebrity video', 'Altered political speech', 'Manipulated news image', 'False social media post']
trust_threshold = 3
# Create deepfake_examples list using a for loop and if condition
# Your code here

Hint

Use range(len(examples)) to loop over indexes and add items with index less than trust_threshold to deepfake_examples.

4

Add advice on verifying information

Create a variable called advice and set it to the string 'Always check multiple sources before trusting online content.'

AI for Everyone

examples = ['Fake celebrity video', 'Altered political speech', 'Manipulated news image', 'False social media post']
trust_threshold = 3
deepfake_examples = []
for i in range(len(examples)):
    if i < trust_threshold:
        deepfake_examples.append(examples[i])
# Add advice string variable
# Your code here

Hint

Assign the exact advice string to the variable advice.

Practice

1. What is a deepfake?

easy

A. A software update

B. A type of computer virus

C. A fake video or image made by AI that looks real

D. A real news report

Solution

1. Step 1: Understand the definition of deepfake

   A deepfake is created using AI to make fake videos or images that look real.

2. Step 2: Compare options with the definition

   Only a fake video or image made by AI that looks real matches this definition exactly.

3. Final Answer:

   A fake video or image made by AI that looks real -> Option C

4. Quick Check:

   Deepfake = Fake AI video/image [OK]

Hint: Deepfake means fake AI-made video or image [OK]

Common Mistakes:

• Confusing deepfake with virus
• Thinking deepfake is real news
• Mixing deepfake with software updates

2. Which of these is a common way to spot a deepfake?

easy

A. Check for unusual blinking or facial movements

B. Look for spelling mistakes in the video

C. Check the video length

D. See if the video is in black and white

Solution

1. Step 1: Identify signs of deepfakes

   Deepfakes often have unnatural blinking or strange facial movements because AI struggles with these details.

2. Step 2: Evaluate options

   Only Check for unusual blinking or facial movements describes a known sign of deepfakes; others are unrelated.

3. Final Answer:

   Check for unusual blinking or facial movements -> Option A

4. Quick Check:

   Unnatural blinking = Deepfake clue [OK]

Hint: Look for weird blinking or face moves [OK]

Common Mistakes:

• Thinking spelling mistakes appear in videos
• Checking video length as a clue
• Assuming black and white means fake

3. If a deepfake video shows a politician saying something shocking, what should you do first?

medium

A. Check trusted news sources to verify

B. Share it immediately on social media

C. Ignore it completely

D. Assume it is true because it looks real

Solution

1. Step 1: Understand the risk of misinformation

   Deepfakes can spread false information, so verifying facts is important.

2. Step 2: Choose the best action

   Checking trusted news sources helps confirm if the video is real or fake.

3. Final Answer:

   Check trusted news sources to verify -> Option A

4. Quick Check:

   Verify before sharing = Check trusted news sources to verify [OK]

Hint: Always verify shocking videos with trusted news [OK]

Common Mistakes:

• Sharing without checking
• Ignoring important info
• Assuming video truth from appearance

4. A video looks real but has strange lighting and mismatched shadows. What is likely the problem?

medium

A. The video is corrupted

B. It is a deepfake with editing errors

C. The camera lens is dirty

D. The video is in slow motion

Solution

1. Step 1: Analyze visual clues in the video

   Strange lighting and mismatched shadows often happen in deepfakes due to AI editing limits.

2. Step 2: Match clues to options

   It is a deepfake with editing errors explains these signs as deepfake editing errors; others do not fit the symptoms.

3. Final Answer:

   It is a deepfake with editing errors -> Option B

4. Quick Check:

   Lighting/shadow errors = Deepfake sign [OK]

Hint: Odd shadows often mean deepfake editing errors [OK]

Common Mistakes:

• Thinking video corruption causes shadows
• Blaming dirty lens for lighting
• Confusing slow motion with lighting issues

5. You receive a video showing a celebrity endorsing a product, but you suspect it is a deepfake. What is the best way to confirm its authenticity?

hard

A. Share the video to see public reaction

B. Trust the video because it looks professional

C. Ask friends if they believe the video

D. Search for official statements or videos from the celebrity

Solution

1. Step 1: Understand how to verify suspicious videos

   Checking official sources or statements is the most reliable way to confirm if a video is real.

2. Step 2: Evaluate options for confirming authenticity

   Search for official statements or videos from the celebrity uses trusted official info; others rely on appearance or opinions, which can be misleading.

3. Final Answer:

   Search for official statements or videos from the celebrity -> Option D

4. Quick Check:

   Official sources confirm authenticity [OK]

Hint: Verify with official celebrity sources first [OK]

Common Mistakes:

• Trusting video quality alone
• Relying on friends' opinions
• Sharing before verifying