Misinformation Detection Framework
Provides a systematic checklist to evaluate claims, identify logical fallacies, assess source credibility, and reach a confidence-rated verdict on content accuracy.
Category: learning
Difficulty: intermediate
Platforms: chatgpt claude
Tags: misinformation fact-checking critical-thinking media-literacy information-literacy
Prompt Template
You are an information literacy expert and fact-checking analyst. Evaluate the following content for potential misinformation.
## Content Under Review
Content type: {{content_type: article/post/video/claim}}
Source: {{source}}
Content or claim to evaluate: {{content_to_evaluate}}
## 1. Red Flag Checklist
Score each indicator (0 = absent, 1 = mild, 2 = strong):
| Red Flag | Score | Evidence |
|----------|-------|----------|
| Sensationalist language or ALL-CAPS | | |
| Missing or anonymous attribution | | |
| No dates or outdated references | | |
| Emotional manipulation over facts | | |
| Cherry-picked statistics | | |
| Lack of primary sources cited | | |
| Claims of secret knowledge or suppression | | |
| Urgency pressure ("share before deleted!") | | |
Total red flag score: __ / 16
## 2. Logical Fallacy Check
Identify any fallacies present:
- Ad hominem, straw man, false dichotomy, appeal to authority, slippery slope
- Correlation vs. causation errors
- Hasty generalization or anecdotal evidence used as proof
- For each fallacy found, quote the exact passage and explain why it qualifies
## 3. Source Credibility Matrix
| Criterion | Rating (1-5) | Notes |
|-----------|-------------|-------|
| Track record of accuracy | | |
| Transparency (authors, funding, corrections) | | |
| Editorial process (peer review, editors) | | |
| Domain expertise match | | |
| Independence from conflict of interest | | |
Credibility score: __ / 25
## 4. Verdict
- **Accuracy assessment**: Confirmed / Likely accurate / Uncertain / Likely inaccurate / Confirmed false
- **Confidence level**: High / Medium / Low (explain why)
- **Key evidence** supporting or refuting the claim
- **Recommended actions**: what the reader should verify independently
- **Better sources** to consult on this topic
Tips
- Paste the exact text of the claim — paraphrasing can alter what is being evaluated
- Always cross-reference the AI verdict with at least one independent fact-checking site
- Use this framework as a thinking scaffold, not a final authority
- The red flag score is directional: a high score warrants deeper investigation, not automatic rejection
- Apply this to your own beliefs too — confirmation bias is the hardest to catch