SYSTEMshift

2025

Omas Gegen Rechts - Chatbot

ai-chatbot anti-propaganda ux-strategy activism ethical-ai civil-courage

SYSTEMshift

2025

Omas Gegen Rechts - Chatbot

ai-chatbot anti-propaganda ux-strategy activism ethical-ai civil-courage

About the Project

The Omas Gegen Rechts movement – a grassroots network founded in Austria and now active in Germany with over 8,000 official members and more than 35,000 supporters across around 200 regional groups – needed a digital companion to help citizens recognize extremist rhetoric, respond safely, and strengthen solidarity. Known for their public demonstrations, educational work, and visible presence at rallies, the movement unites older women and allies who stand clearly against right‑wing extremism and for democracy. We built a chatbot persona – Klara Klartext – programmed with guardrails, fact-based debunking strategies, and conversational coaching to empower people in everyday contexts, from online debates to the local Stammtisch.

Challenge

Right-wing propaganda spreads quickly online and in person. Many citizens want to speak up but lack safe, fact-based ways to respond. Activists also need scalable tools for training and fieldwork.

Solution

We designed a collective persona – Klara Klartext. The chatbot offers empathy, clarity, and concrete options. It integrates psychological techniques like motivational interviewing and prebunking, combined with fact-checking and historical context.

Impact

The chatbot helps ordinary citizens feel prepared to counter hate, activists gain new training capacity, and the Omas Gegen Rechts movement extend their reach digitally.

Omas Gegen Rechts - Chatbot

How it works

  • Welcome: Greets users warmly, sets a tone of solidarity.

  • Modes: Choose AfD program analysis, Train-the-Trainer, or Everyday Empowerment.

  • Formats: Mythos vs. Fakt, Kurz gekontert, Aktionsplan, Erkennungszeichen, etc.

  • Guidance: Always ends with next-step prompts so users stay engaged and build confidence.

  • Guardrails: Refuses extremist reproduction, maintains fact-based, empathetic tone.

Safety and trust

  • Clear disclaimers: not legal or psychological advice.

  • Encourages referral to helplines/authorities if danger is present.

  • Privacy-first: no long-term chat storage.

Sponsor Users

Grassroots activists, educators, and citizens across Germany seeking empowerment.

Product Design Team

Activist volunteers, movement organizers, design collaborators

My Contributions

Persona definition · Prompt/guardrail design · UX copywriting · Legal/ethical framework · MVP implementation · Stakeholder facilitation · App Development

Metrics

  • 3 engagement modes: AfD program decoding, activist training, and everyday citizen empowerment

  • 6+ standardized response formats (Mythos vs. Fakt, Kurz gekontert, Aktionsplan, etc.)

  • Safety guardrails: no propaganda reproduction, no doxing, no hate amplification

Tools

  • OpenAI

  • Replit

  • React/Tailwind

  • Custom prompt/guardrail framework