1. Which of these is something AI tools currently CANNOT reliably do in the fact-checking process?
Correct. Live lateral reading — searching what credible sources currently say about a specific outlet — requires real-time web access and human judgment that standard AI tools cannot provide.
AI cannot browse the web in real time to check what credible sources currently say about a specific outlet — that requires the human to do lateral searching themselves.
2. The Google prebunking campaign in Poland, Czech Republic, and Slovakia (2022–2023) used what format to deliver inoculation content?
Correct. Short pre-roll YouTube ads explaining manipulation techniques — false urgency, scapegoating — without naming specific content reached 90+ million impressions and showed measurable effect in research.
The campaign used short YouTube pre-roll ads explaining manipulation techniques (false urgency, appeals to inconsistency) without mentioning specific false content — reaching over 90 million impressions across the three countries.
3. The SAG-AFTRA 2023 AI settlement established that studios must obtain what before using AI to replicate a performer?
Correct! The SAG-AFTRA settlement required both consent AND compensation — establishing that a performer's likeness is their property and cannot be AI-replicated without their informed agreement and fair pay.
The settlement required explicit informed consent plus negotiated compensation — both elements, not just one. This established the ethical framework: consent is necessary but not sufficient without fair compensation.
4. In the three-layer amplification model, which layer involves state-affiliated media resharing content before fact-checkers can respond?
Correct. Layer 2 (Relay) is where accounts with genuine large followings or state media infrastructure amplify content, adding perceived legitimacy through source laundering before correction arrives.
Layer 2 (Relay) describes state-affiliated or high-follower accounts resharing content — adding legitimacy through source laundering and outpacing the fact-checking timeline.
5. What was the Meta Oversight Board's ruling on Meta's indefinite ban of Trump's account?
Correct. The Board upheld that action was warranted but found the open-ended nature of the ban wasn't within Meta's published rule set — illustrating that procedural integrity matters even when the substantive outcome is right.
The Oversight Board found the content warranted action but that an indefinite ban was inconsistent with Meta's rules — the right outcome through the wrong process is still a procedural failure.
6. How did the Russian IRA evade platform behavioral detection systems, per the Senate Intelligence Committee report?
Correct. The IRA used slow account building, organic-seeming content mixed with influence posts, and controlled velocity — adversarial calibration against the known behavior of platform detection systems.
The IRA used adversarial evasion: slow account building, mixing authentic content with influence posts, and carefully limiting posting velocity to avoid spam triggers.
7. What is "surface fluency" in the context of AI-generated misinformation?
Correct. Surface fluency refers to AI's ability to produce grammatically correct, well-organized text by statistical pattern-matching — without any connection between writing quality and factual accuracy. This breaks the longstanding heuristic readers use to evaluate sources.
Surface fluency means AI produces polished, credible-sounding prose purely through pattern-matching — with no connection to accuracy. This breaks the "good writing = reliable source" heuristic humans have used for generations.
8. What was the accuracy rate of OpenAI's AI Text Classifier for detecting AI-written text before it was shut down?
Correct. The classifier correctly identified only 26% of AI-written text — meaning it missed roughly three-quarters of what it was designed to catch — while generating false positives on 9% of genuinely human-written text. OpenAI determined this accuracy level was insufficient and potentially harmful.
The classifier caught only 26% of AI-written content — missing three out of four AI-generated texts — while falsely flagging 9% of human-written text. This was OpenAI's own assessment, which led to the tool being shut down just six months after launch.
9. The EU AI Act's requirement regarding deepfakes is that:
Correct! The EU AI Act requires transparency — deepfakes must be labeled as AI-generated. Violations can result in fines of up to 3% of global annual turnover. The rules take full effect from 2025.
The EU AI Act mandates labeling — synthetic and AI-generated content must be disclosed as such. It doesn't ban deepfakes outright but requires transparency and platform accountability, with significant financial penalties.
10. The 2022 Zelensky deepfake was posted on:
Correct! The Zelensky deepfake was placed on hacked Ukrainian media sites to give it false credibility, then spread rapidly via Telegram and Facebook before platforms removed it.
The deepfake was injected into hacked Ukrainian TV and news sites — giving it apparent credibility as "official" Ukrainian media — and then spread across Telegram and Facebook.
11. Lateral reading is best described as:
Correct. Lateral reading means moving across the web to find what credible third parties say about a source.
Lateral reading means leaving a page quickly to search what others say about it — not reading it more carefully.
12. Which tool is specifically useful for finding the OLDEST appearance of an image online?
Correct. TinEye's "sort by oldest" feature is the primary tool for proving that an image predates the event it supposedly depicts.
TinEye with the "oldest" sort filter is the right tool for this specific task.
13. In May 2021, Facebook reversed its policy of removing lab leak hypothesis posts. The reversal acknowledged that the hypothesis was a:
Correct. The reversal acknowledged the hypothesis was contested, not settled — the distinction at the heart of the Lesson 3 taxonomy. It became a defining case for why platforms should not treat contested empirical claims as misinformation.
Facebook acknowledged the lab leak hypothesis was a contested empirical claim — not settled misinformation. Treating contested claims as settled is the epistemic humility failure the lesson identifies.
14. How did the NewsGuard investigation illustrate the "flood strategy" in AI misinformation?
Correct. NewsGuard found sites growing from 49 to over 700 in four months, each publishing hundreds of articles per day. The economic implication is stark: AI removed the human labor constraint that previously limited the scale of misinformation operations.
NewsGuard's investigation found AI-generated fake news sites publishing at massive volume — growing from 49 to over 700 sites in four months. This demonstrates how AI eliminates the human labor ceiling that previously constrained the scale of misinformation production.
15. The Bellingcat principle described in Lesson 4 is:
Correct. Investigators like Bellingcat and EU DisinfoLab focus on infrastructure — registration records, hosting, IP addresses — because that is what misinformation creators neglect to disguise.
The principle is that content is carefully designed to deceive, but infrastructure (WHOIS, hosting, IP) is often neglected — so that's where investigators look.
16. "Citation laundering" is best countered by which SIFT move?
Correct. Tracing the citation chain is the only way to reverse the laundering process and find out what the original source actually said.
Citation laundering hides weak sources inside chains of citations. Only "Trace" — following that chain back to the primary document — can expose it.
17. Which content moderation technique does NOT involve machine learning probabilistic judgment?
Correct. Hash matching compares exact digital fingerprints — a match is a match, with no probabilistic judgment involved.
Hash matching uses exact digital fingerprint comparison — no ML judgment. A match is definitive, which is why it's the most reliable automated technique.
18. According to the five-question decision framework, what is the role AI should play in content moderation decisions on ambiguous cases?
Correct. The framework positions AI as a triage and confidence-scoring system for surfacing content and relevant policy — with human reviewers making final decisions on ambiguous cases where context, editorial significance, or harm asymmetry are at stake.
Per the framework, AI should triage, flag, and confidence-score — but humans should make final decisions on ambiguous cases. Context assessment, editorial significance, and harm asymmetry require human judgment.
19. WHOIS records are primarily used by investigators to discover:
Correct. WHOIS records reveal registration dates and infrastructure details — crucial for spotting networks of coordinated fake sites.
WHOIS reveals registration details: dates, registrants (when not hidden), and shared infrastructure that can link seemingly independent sites.
20. According to AFP Fact Check's published methodology, what verification step is required before any image is accepted as evidence of a specific event?
Correct. AFP's methodology explicitly requires reverse image verification before any image can be accepted as evidence of the event it supposedly depicts.
AFP's standard is reverse image verification — no image can be accepted as evidence without it.