1. AI as a lever for climate justice

The climate crisis hits marginalized communities hardest. But AI can help here:

  • Tools such as WattTime used satellite data and machine learning to track 352 million global emissions sources in real time – a technology that has also revolutionised Munich’s waste management.

Algorithms are neutral tools. But who controls them decides whom they serve.

2. Housing for all: Generative AI in social housing

The housing crisis affects low-income earners: women particularly strongly. Generative AI offers solutions:

  • Projects such as Agrimetrics analyze land prices and building permits to develop low-cost housing concepts. In Berlin, we are testing similar models for social housing.

  • A Munich pilot project uses AI to identify vacancies in luxury properties and convert them into affordable housing – an approach that gave 200 families a home in 2025.

3. Activism 2.0: How AI Strengthens Protests

  • FactStream: This tool checks politicians’ inside statements in real time – used in the EU election campaign 2024 to expose misinformation.

  • Geneva: AI-based software that circumvents censorship in authoritarian regimes. She helped us establish secure channels of communication during the protests in Belarus.

  • Debunk.eu: A European project that uses NLP (Natural Language Processing) to uncover disinformation networks – a model that we have adapted for climate deniers.

4. Risks: When Algorithms Cement Inequality

AI is not a panacea. Without critical control, it reproduces power structures:

  • Bias in social welfare: An OECD study shows that AI systems often reinforce racist and gender bias when granting minimum wage applications.

  • Predictive Policing: In Munich in 2023, we blocked the use of an algorithm that labeled districts with a high proportion of migrants as “crime hotspots”.

Formal fairness is not enough. AI must be substantially just – or it becomes a tool of oppression. (cf. Harvard study on algorithmic fairness)

5. Solutions: How AI becomes a commons

a. Open source for everyone

Platforms such as Hugging Face or BLOOM enable communities to use AI models without budget limits. My tip: Local hackathons where activists and developers build tools for concrete problems – such as our “climate dashboard” for Munich schools.

b. Ethics framework with teeth

  • Transparent data sets: RECUP only uses AI when training data is publicly available.

  • Civil society control bodies: In our alliance, we call for AI audits by independent NGOs.

c. Education as a Key

Workshops such as “AI for Activists” provide basic knowledge – such as how to use algorithms for petitions or optimize social media campaigns.

6. My appeal: AI must be thought activistically

AI is not a substitute for street protests – but it can amplify them. When I organized the “Sea of Lights for Democracy” demo in 2024, algorithms helped us to plan routes and coordinate participants safely. At the same time, we are fighting tech companies that use AI for surveillance.

The question is not whether AI is powerful. The question is: Who controls them – and for whom?