Deepfakes and misinformation in AI for Everyone - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

What are deepfakes primarily used for?

Deepfakes are synthetic media where a person's likeness is replaced or altered using artificial intelligence. What is the main purpose of creating deepfakes?

A To automatically translate spoken language in videos

B To create realistic fake videos or images that can mislead viewers

C To enhance the resolution of old photographs

D To improve video compression for faster streaming

💡 Hint

Think about why someone would want to convincingly change what a video or image shows.

Deepfakes are mainly created to produce realistic but fake videos or images that can deceive people by showing events or statements that never happened.

📋 Factual

intermediate

Which technology is most commonly used to create deepfakes?

Deepfakes rely on a specific type of artificial intelligence technology to generate realistic fake media. Which technology is it?

A Generative Adversarial Networks (GANs)

B Decision Trees

C Support Vector Machines (SVM)

D K-Nearest Neighbors (KNN)

💡 Hint

This technology involves two neural networks competing to improve the quality of generated images or videos.

Generative Adversarial Networks (GANs) are the AI models most commonly used to create deepfakes by generating realistic images or videos through a competition between two networks.

🔍 Analysis

advanced

What is a major risk of deepfakes in misinformation campaigns?

Consider how deepfakes can be used in spreading false information. What is a key risk they pose in misinformation campaigns?

A They automatically detect and remove fake news from social media

B They reduce the quality of real videos making them hard to watch

C They make it impossible to create any real videos in the future

D They can create believable fake evidence that damages reputations or influences opinions

💡 Hint

Think about how a fake video might affect public trust or someone's image.

Deepfakes can produce convincing fake videos that appear real, which can be used to falsely accuse or discredit people, or manipulate public opinion.

❓ Comparison

advanced

How do deepfakes differ from traditional photo or video editing?

Compare deepfakes with traditional editing methods. What is a key difference?

A Deepfakes are easier to detect than traditional edits because they leave visible marks

B Traditional editing creates fake content, but deepfakes only enhance real footage

C Deepfakes use AI to generate entirely new realistic content, while traditional editing modifies existing content manually

D Traditional editing requires AI, but deepfakes do not

💡 Hint

Think about how deepfakes generate new images versus how traditional editing changes existing ones.

Deepfakes generate new, realistic images or videos using AI, often creating content that never existed, while traditional editing changes or removes parts of existing media manually.

❓ Reasoning

expert

Why is it challenging to combat misinformation caused by deepfakes?

Given the nature of deepfakes, why is it difficult for society to effectively stop misinformation spread by them?

A Because deepfakes can be very realistic and spread quickly before detection tools can respond

B Because deepfakes are always low quality and easy to spot, so they are ignored

C Because deepfakes require expensive equipment that only governments have access to

D Because deepfakes only appear in private messages and never on public platforms

💡 Hint

Consider how fast information spreads online and how convincing deepfakes can be.

Deepfakes are often very realistic and can be shared widely and quickly on social media, making it hard for fact-checkers and detection tools to keep up and prevent misinformation from spreading.