The narrative shift toward AI autonomy has already happened — we just haven't noticed. We analyzed 10,500 headlines to find out how.
Everything you need to explore our research — code, presentation, profiles, and documentation.
Full reproducible analysis. Run all cells — 8–12 min on CPU. No API keys required.
Full slide deck with all findings, visualizations, and AI prompt appendix.
You're here. Full findings with live charts, interactive bias network, and all metrics.
CS & Data Science (Statistics). Minor in Cognitive Science & Statistics.
Data Science (CS Track). Minor in Business Administration.
10,500 AI news headlines (Jun–Sep 2025). Available to competition judges only.
Others analyze what media says. We analyze how it distorts thinking. Using NLP, cognitive psychology, and four original metrics, we quantify psychological manipulation in AI journalism.
8 types of psychological biases in AI headlines — mental shortcuts that distort how readers understand AI.
Quantifies who's framed as in control: humans or AI. PSI > 100 = AI framing dominates.
Composite 6-dimension fear measure: negative sentiment, fear keywords, risk words, economic threat, health/safety, ethics signals.
Tracks deep human experiences — love, grief, spirituality, community — that AI coverage systematically ignores.
News doesn't neutrally report on AI — it activates psychological triggers. 53.9% of headlines stack multiple biases simultaneously.
💡 Hover over nodes and connections for details · Drag to rearrange
The narrative shift isn't a future trend. It's already embedded in current coverage — for every headline showing human control, there are 1.45 showing AI agency.
AI coverage doesn't create mass panic — but it reliably channels fear toward specific, concentrated themes. The formula is: Jobs + Fear language + Loss of Control = Maximum Anxiety.
| Predictor | r with AAI | Visual |
|---|---|---|
| Fear Bias | 0.456 | |
| Economic Threat | 0.285 | |
| Moral Panic | 0.249 | |
| Bias Intensity | 0.229 | |
| Control Loss | 0.173 | |
| Techno-Utopianism | 0.131 | |
| Optimism Bias | 0.097 | |
| Anthropomorphism | 0.074 |
The Formula: Jobs + Fear Language + Loss of Control = Maximum Anxiety
High-Anxiety Headlines
Low-Anxiety Headlines
Media reduces AI to Work + Money + Efficiency, systematically ignoring how AI affects love, loss, faith, community, and what it means to be human.
All models are open-weights. No proprietary AI. Fully reproducible with seed=42.
Uses sentence-transformers/all-MiniLM-L6-v2 for semantic similarity to 8 prototype bias descriptions. Custom 84-keyword lexicons per bias category. VADER sentiment analysis as supporting signal.
Adapted from computational media studies agency-framing research. PSI = 100 is balance; above 100 means AI agency framing dominates. Keyword sets validated via semantic embeddings.
NS=Negative sentiment, FI=Fear intensity (highest weight: 3.5), RI=Risk intensity, ET=Economic threat, HS=Health/safety, ES=Ethics signals. Composite 6-dimension weighted score.
Keyword-based domain detection across 8 human experience categories: artistic process, ethics, work/employment, community, intimacy, grief, embodiment, spirituality. Validated with semantic similarity scoring.
sciPy Pearson r, p-values. At n = 10,500, findings are not due to chance. Bias intensity ↔ AAI: r=0.64. Economic threat ↔ work topics: r=0.78. Techno-utopianism ↔ control loss: r=0.85 (strongest).
Also: numpy, pandas, matplotlib, scipy, MinMaxScaler. All models open-weights. Random seeds set (seed=42). Single-click execution via "Runtime → Run All." Runtime: 8–12 min on CPU.
Biased AI coverage doesn't just shape individual opinions — it reverberates across society, policy, and industry.
Grounded in data. Targeting the specific mechanisms that create distorted AI coverage.
Note on Credibility Scores: Bias intensity varies 2–3× across different news sources in our dataset. Our Credibility Score (Score = 100 − Bias Intensity × 100) ranged from 35 to 65 across outlets — outlet choice matters significantly for balanced AI information.