A TikTok video goes viral — a politician says something outrageous. Millions share it. But it never happened. The clip was generated by AI in under a minute. In the previous article, you learned that AI can hallucinate — produce wrong information by accident. This article goes one step further: what happens when people deliberately use AI to deceive? You will learn what deepfakes are, how to quickly verify sources, and how to navigate a world where most content may be AI-generated.

Deepfakes — When AI Lies

Think of a deepfake like a Hollywood-quality costume and mask — but digital. In the past, creating a convincing disguise required a professional makeup team and hours of work. Now AI provides the mask, voice, and body language instantly — and the result fools even people who know the real person. The mask is so perfect that you cannot spot the seams. Breaking point: A real costume can be touched and removed — a deepfake exists only as digital data.

Analogy:

Think of a deepfake like a Hollywood-quality costume and mask — but digital. In the past, creating a convincing disguise required a professional makeup team and hours of work. Now AI provides the mask, voice, and body language instantly — and the result fools even people who know the real person. The mask is so perfect that you cannot spot the seams. Breaking point: A real costume can be touched and removed — a deepfake exists only as digital data.

Definition:

A deepfake is media (video, image, or audio) generated or manipulated by AI to show a person doing or saying something that never happened. Video deepfakes swap faces. Audio deepfakes (voice cloning) replicate a voice from as little as 3 to 15 seconds of sample audio. Image deepfakes generate entirely fictional people or place real people in fabricated situations. By 2025, professional deepfakes are virtually undetectable by human perception.

Deepfakes ~2019

Visible artifacts: wrong finger count, unnatural blinking, blurry facial edges, inconsistent lighting. Detectable by attentive observers.

Deepfakes 2025

Professional quality: correct anatomy, natural facial expressions, consistent lighting, lip-synced speech. Virtually indistinguishable from real footage by the human eye.

An account called @deeptomcruise posted videos of "Tom Cruise" playing golf and performing magic tricks. The videos were so convincing that millions of viewers believed they were real. Only after investigation was it revealed that a professional deepfake artist created them. The lesson: if even a globally recognizable face can be faked convincingly, any face can be faked.

That may have been true in 2019 — weird fingers, strange blinking, blurry edges. But not anymore. Modern deepfakes have largely eliminated these artifacts. Relying on visual detection gives you false confidence. The right approach: verify the source, not the pixels.

Deepfakes are frequently based on Generative Adversarial Networks (GANs): two neural networks train each other — a generator creates fakes, a discriminator tries to detect them. Through this interplay, the fakes keep improving. In video deepfakes, a face is mapped onto another face frame by frame. In voice cloning, AI analyzes speech patterns, pitch, and rhythm from a short audio sample and then synthesizes any statement in that voice.

SIFT — Your Tool Against Fakes

Imagine you are a detective following up on a tip. A detective never takes a single witness statement at face value. They check: Who is this witness? (Investigate) Are there other witnesses who confirm the story? (Find) Can I trace the claim back to hard evidence? (Trace) And the first rule of any investigation: do not jump to conclusions (Stop). SIFT makes you the detective of your news feed. Breaking point: A detective has legal authority — you rely on publicly available information.

Analogy:

Imagine you are a detective following up on a tip. A detective never takes a single witness statement at face value. They check: Who is this witness? (Investigate) Are there other witnesses who confirm the story? (Find) Can I trace the claim back to hard evidence? (Trace) And the first rule of any investigation: do not jump to conclusions (Stop). SIFT makes you the detective of your news feed. Breaking point: A detective has legal authority — you rely on publicly available information.

Definition:

SIFT is a four-step source verification method developed by media literacy expert Mike Caulfield (2019). The four steps: Stop (do not react immediately), Investigate (check the source), Find (look for other coverage), Trace (follow the claim back to its primary source). SIFT makes you the detective of your own news feed.

1

Stop Do not react or share immediately. Disinformation exploits emotional reactions.

2

Investigate Check the source: Who published this? Reputable outlet or anonymous account?

3

Find Look for other coverage: Do independent, reputable sources report the same thing?

4

Trace Follow the claim to its primary source: Which study? By whom? Does it exist?

A viral post claims: "Scientist confirms: [shocking health claim]." 50,000 shares. SIFT check: **Stop** — Do not share. **Investigate** — The poster has 200 followers, no credentials, account created 3 days ago. **Find** — No reputable health outlet, medical journal, or news agency reports this. **Trace** — The cited "study" does not exist: no DOI, no author, no journal. Result: fake. Time spent: about 90 seconds.

The SIFT method is designed for speed: steps 1 and 2 take under 30 seconds. You do not need to check every post — only those that trigger strong emotions, make extraordinary claims, or ask you to act (share, click, donate). Disinformation thrives on urgency — pausing for just 10 seconds breaks its power.

Interactive: Check a Source Yourself

You have learned the SIFT method. Now apply it: work through the checklist and evaluate any online source. At the end you will receive a credibility score with a traffic-light rating — so you can see at a glance how trustworthy the source is.

Credibility Check

Check an online source for trustworthiness. Answer the following questions with Yes or No.

1.Is the author named and verifiable?

Anonymous sources are less trustworthy. Check via LinkedIn, university websites, etc.

2.Does the source have an imprint or an "About us" page?

Reputable media identify themselves. Missing information is a warning sign.

3.Do other independent sources report the same thing?

SIFT step "Find": At least 2 independent sources should confirm the core claim.

4.Are specific studies or data cited?

Vague claims like "studies show..." without citation are a warning sign.

5.Are the cited sources verifiable and existing?

SIFT step "Trace": Does the cited study exist? Does it really say what is claimed?

6.Is the language factual and free from emotionalization?

Extreme phrasing ("SHOCKING!", "The truth THEY are hiding") indicates manipulation.

7.Is the publication date current and visible?

Old information without date context can be misleading.

8.Are there no obvious deepfake or AI manipulation signs?

For images: Reverse Image Search. For videos: Check source. Look for AI markers.

0 / 8 answered

Synthetic Media — The Big Picture

Synthetic media is the umbrella term for all content — text, images, video, audio — that is wholly or partially generated by AI. This includes deepfakes but also useful applications like AI translations, automatic subtitles, image enhancement, and AI art. According to the Europol report "Facing Reality," an estimated 90 percent of online content could be synthetically generated or enhanced. The core problem is not synthetic media itself but the lack of labeling.

3-15 sec

Voice Cloning That is how little audio material AI needs to convincingly clone a voice

~90%

Synthetic Content Projected share of AI-generated or AI-enhanced online content (Europol expert forecast)

30 sec

SIFT Quick Check That is how long the first two SIFT steps (Stop + Investigate) take — faster than a TikTok video

Reverse image search: (Google Lens) — Has this image appeared before in a different context?
Check metadata: AI-generated images often lack EXIF data from a real camera.
Look for C2PA labels: Some platforms already show "AI-generated" badges.
Apply SIFT: Does a credible source confirm the content?

If all four checks fail to confirm authenticity: treat the content as unverified.

Only partially true. Detectors like GPTZero or Hive are useful but have significant error rates and are in a constant arms race with generators. Generators are always one step ahead because generating is computationally easier than detecting. Detectors are a helpful addition, but not a definitive answer. Your SIFT skills remain essential.

Article 50 of the EU AI Act requires that AI-generated content must be labeled as such — a first regulatory step against invisible synthetic media. In parallel, the C2PA consortium (Content Provenance and Authenticity) is developing a technical standard: a digital provenance certificate embedded in media files that shows when, where, and with what tool content was created or modified. Some platforms already support C2PA.

Deepfakes have become too good: Video, audio, and image deepfakes can no longer be reliably detected by human senses. "I can see it's fake" is a dangerous assumption.
SIFT is your universal tool: Stop → Investigate → Find → Trace. These four steps work for any type of online content — deepfake videos, AI-generated texts, or manipulated images.
Synthetic media are not inherently bad: AI translations, accessibility features, and art are useful. But missing labeling and malicious use make synthetic media dangerous. Your verification skills are the last line of defense.

This is the final article in the curriculum. The SIFT method goes far beyond deepfakes — it is your tool for evaluating any information, in any medium, for the rest of your life.

What is a deepfake?

A virus that infects your computer through fake downloads

AI-generated or AI-manipulated media that shows a person doing or saying something that never happened

A social media filter that makes photos look better

An encrypted message that cannot be read by others

1. What is a deepfake?

☐ A) A virus that infects your computer through fake downloads
☐ B) AI-generated or AI-manipulated media that shows a person doing or saying something that never happened
☐ C) A social media filter that makes photos look better
☐ D) An encrypted message that cannot be read by others

2. A friend sends you a shocking video of a celebrity making a controversial statement. Applying the SIFT method, what should you do FIRST?

☐ A) Share it immediately so others can see it too
☐ B) Check the video for visual glitches like wrong finger count
☐ C) Stop — do not react or share immediately
☐ D) Report the video to the platform

3. You see a news post: "New study proves: [shocking health claim]." 50,000 shares. You investigate: anonymous account, no major outlet reports it, the "study" has no author and no DOI. What is your conclusion?

☐ A) It must be true because 50,000 people shared it
☐ B) The claim is unverified and should not be trusted or shared — anonymous source, no independent coverage, study cannot be traced
☐ C) It is definitely true because it mentions a study
☐ D) You need to wait for an AI detector to confirm whether it is real

4. A classmate argues: "We don't need to learn source verification — AI detectors like GPTZero will identify all fakes for us." Why is this argument flawed?

☐ A) GPTZero is too expensive for students
☐ B) AI detectors have significant error rates and are in a constant arms race with generators — detectors are a helpful addition but not a replacement for SIFT-based verification skills
☐ C) AI detectors only work for text, not for images or videos
☐ D) AI detectors require a monthly subscription that schools cannot afford

Answer Key: 1) B · 2) C · 3) B · 4) B

Learning Goals

What are the three types of deepfakes and why does visual detection no longer work?
What are the four SIFT steps in the correct order and how would you apply them to a specific example?
Why are AI detectors alone not enough and which tool remains essential?

Truth Online

Deepfakes — When AI Lies

Deepfake

Analogy:

Definition:

Misconception: "I can spot deepfakes — there are always glitches"

Deep Dive: How Deepfakes Work Technically

SIFT — Your Tool Against Fakes

SIFT Method

Analogy:

Definition:

Walkthrough: Fake Health Post on TikTok

Misconception: "Source verification takes too long"

Interactive: Check a Source Yourself

Credibility Check

Synthetic Media — The Big Picture

Quick Check: Is This Real?

Misconception: "AI detectors solve the problem"

Deep Dive: EU AI Act & C2PA

Key Takeaways

What This Is About

What is a deepfake?

Learning Goals

Deepfakes — When AI Lies

Deepfake

Analogy:

Definition:

Misconception: "I can spot deepfakes — there are always glitches"

Deep Dive: How Deepfakes Work Technically

SIFT — Your Tool Against Fakes

SIFT Method

Analogy:

Definition:

Walkthrough: Fake Health Post on TikTok

Misconception: "Source verification takes too long"

Interactive: Check a Source Yourself

Credibility Check

Synthetic Media — The Big Picture

Quick Check: Is This Real?

Misconception: "AI detectors solve the problem"

Deep Dive: EU AI Act & C2PA

Key Takeaways

What This Is About

What is a deepfake?

Learning Goals

Related Content

Article

Safety & Fraud Protection

A Digital Helper — AI in Everyday Life

Learning with AI (Instead of Cheating)

Staying in Control

Demo

ELIZA

GANs (Generative Models)

Glossary

Timeline