How It Works

Understand the analysis pipeline, frameworks, and limitations of PSYOP Detector.

What is PSYOP Detector?

PSYOP Detector is an experimental tool that analyzes news articles for signs of psychological manipulation, propaganda techniques, and narrative framing. When you submit a URL, we extract the article text and run it through three independent analysis passes using a large language model.

The Three-Lens Approach

Each article is analyzed through three complementary frameworks to provide a comprehensive picture:

FATE Model

Developed by Chase Hughes, a retired US Navy Chief and behavioral intelligence expert, the FATE model measures four dimensions of psychological manipulation: Focus (how content captures attention through novelty or urgency), Authority (how credibility is leveraged or manufactured), Tribe (how group identity and us-vs-them dynamics are exploited), and Emotion (how fear, outrage, or moral superiority are engineered). Each dimension is scored 0-10 based on specific signals found in the text.

PCP Model & SMRP Pattern

Also from Chase Hughes' behavioral intelligence framework, the PCP model maps the psychological pipeline of influence operations: Perception (what beliefs are being targeted), Context (what information is shifted or omitted), and Permission (what behavior the audience is being nudged toward). The SMRP pattern detects maintenance tactics used to sustain manipulation: Socializing ideas as normal, Minimizing concerns, Rationalizing with logic, and Projecting blame onto opponents.

SemEval-2023 Propaganda Techniques

From academic research in computational propaganda detection, this taxonomy identifies 23 specific propaganda techniques across six categories: Manipulative Wording, Attack on Reputation, Justification, Simplification, Call to Action, and Distraction. Each detected technique includes the exact quote from the article and an explanation of how it functions.

The Manipulation Score

The overall manipulation score (0-100) is a weighted composite of three factors: the average FATE dimension scores (40%), propaganda technique density relative to article length (35%), and PCP pipeline completeness including SMRP and red flags (25%). Higher scores indicate more manipulation signals were detected.

Technology

Analysis is performed by LLM analysis. Each article goes through three parallel analysis passes (FATE, PCP, and techniques), followed by a summary generation pass. Results are permanently stored so the same article only needs to be analyzed once.

Disclaimer

  • This is an experimental tool, not a fact-checker. PSYOP Detector identifies patterns commonly associated with manipulation and propaganda. The presence of these patterns does not necessarily mean an article is intentionally deceptive or that its claims are false.
  • AI analysis has limitations. The LLM may produce false positives (flagging legitimate rhetoric as manipulative) or false negatives (missing subtle manipulation). Results should be used as one input in your own critical thinking, not as a definitive judgment.
  • All journalism uses persuasion. Techniques like framing, appeals to authority, and emotional language are present in legitimate reporting as well as propaganda. A high score does not automatically mean an article is "bad" — context matters.
  • Scores are not comparable across topics. Emotionally charged topics (war, politics, social issues) will naturally score higher than neutral reporting regardless of manipulative intent.
  • No political bias is intended. The tool analyzes manipulation techniques regardless of political orientation. Articles from all parts of the political spectrum may score high or low.

Credits

The FATE model, PCP model, and SMRP pattern are based on the work of Chase Hughes, a retired US Navy Chief and behavioral intelligence expert. His frameworks are used by military PSYOPS, intelligence agencies, and security professionals worldwide. The propaganda technique taxonomy is based on the SemEval-2023 Task 3 shared task for computational propaganda detection.

Support PSYOP Detector

PSYOP Detector is an independent project. If you find it useful, consider buying me a coffee to help cover server and AI analysis costs.