TL;DR: The Algorithmic Governance Foundation (AGF) develops and deploys algorithmic solutions to improve public decision-making—especially in low-resource settings. We believe this work addresses a neglected and highly tractable intersection of two core EA cause areas: transformative AI and institutional decision-making. We’re looking for collaborators, co-directors, and volunteers to help grow this work.

Why the AGF Exists

Governments and public-serving institutions make decisions that impact millions of lives. In theory, applying modern algorithmic tools—such as machine learning, simulation modeling, or LLMs—could significantly improve the quality, fairness, and efficiency of those decisions.

Yet, despite rapid advances in AI, most governments are not equipped to harness this potential. Reasons include:

• Decentralised or messy data systems
• Lack of in-house technical expertise
• Limited funding to explore innovative tools
• Complexity in scoping meaningful AI projects

Meanwhile, many public policy challenges—conflict, housing, AMR, climate adaptation—are tractable if decision-makers have better forecasting, risk detection, or data integration tools.

That’s where the Algorithmic Governance Foundation (AGF) comes in.

What We Do

AGF is a new initiative applying algorithmic tools—machine learning, simulations, automations, and large language models—to real-world public policy problems. We partner directly with nonprofits, governments, and public-serving organisations to implement technical solutions that improve decision-making at scale.

With 3-4 volunteers per project, we've completed algorithmic implementations such as:

• Flood Prediction in South Sudan: A neural network-based early warning system (INFLOW-AI) that predicts flooding up to two months in advance. Doctors without borders, the Red Cross, and the UN Refugee Agency used our forecasts to guide aid allocation and vaccination campaigns for approximately 10,000 households. We estimate the system saved humanitarian agencies nearly $800,000 and possibly prevented around 600 cholera cases in the 2024 flood year.
• Crop Yield Forecasting in Bihar: An ensemble machine learning model that doubled the precision of previous yield forecasting systems. Digital Green used the system to better target agricultural support to farmers most likely to benefit, applied to their audience of nearly 700,000 farmers in Bihar. A back-of-the-envelope calculation suggests our tool may have increased farmers' agricultural output by over $3 million during the 2023 harvest.

And we currently working on several other projects:

• Predicting Global Conflict: Enhancing ACLED’s Conflict Alert System using LLM web-scraping (Leximine) to feed in timely news intelligence.
• Hawaii Wildfire Risk Prediction: Deploying ML models to help the Hawaii Wildfire Management Organisation forecast wildfire risks with high spatial resolution.
• Simulating Housing Policy Impacts: Building tools to simulate how housing policies affect affordability—helping governments make evidence-based reforms.
• Monitoring Antimicrobial Resistance (AMR): Scraping global news sources for AMR-related stories in local languages, populating an international AMR risk database.
• Tracking Displaced Populations in Gaza: Using satellite imagery and ML (in collaboration with Forensic Architecture) to monitor displacement and document potential human rights violations.

See more projects at https://algorithmicgovernance.org/.

Our approach is fundamentally different from the many think tanks and advocacy orgs in the "AI for Good" space. Rather than advising others to “do something” about algorithmic governance, we actually build and deploy the solutions ourselves.

We’re Looking for Collaborators

We’re building this as a bottom-up, open-source EA-aligned implementation organisation—and we need help. If you’ve ever thought, “Why aren’t more people actually building AI tools that directly help EA goals?”—this is your invitation to get involved.

Volunteers & Research Leads

We are currently seeking volunteers and research leads for the following groups:

• AI & Data Science Group
  • Useful skills: Data scientists, ML engineers, and statisticians
  • Focus:
    • Develop prediction models, simulations, and automation tools
    • Research ethical AI practices for low-resource settings
    • Work on technical improvements and debugging
• Data Engineering & Infrastructure Group
  • Useful skills: Data engineers, backend developers, and DevOps specialists
  • Focus:
    • Build and maintain data pipelines and cloud infrastructure
    • Ensure scalability and security of AI solutions
    • Optimise data storage and processing for partners
• Policy & Impact Group
  • Research Lead: Policy & Governance Lead
  • Useful skills: Policy analysts, ethicists, and social impact researchers
  • Focus:
    • Ensure compliance with AI ethics and regulatory standards
    • Study the real-world impact of AI models on partner organisations
    • Provide recommendations for responsible AI deployment
• Partnerships & Community Engagement Group
  • Research Lead: Partnerships Lead
  • Useful skills: Outreach coordinators, marketers, and fundraising specialists
  • Focus:
    • Manage relationships with governments, NGOs, and donors
    • Lead volunteer recruitment and onboarding
    • Organise fundraising campaigns and public awareness efforts

Co-Directors

We're also hiring for two leadership roles to help grow AGF:

Director of Development: Focus on fundraising, partnerships (NGOs, governments, tech firms), and community growth. Ideal for someone with experience in EA org development, grant writing, or movement building.

Director of Operations & Impact (optional): Ensure projects are well-implemented, track impact metrics, and handle ops, HR, and compliance. Great fit for someone process-oriented who’s excited about implementation logistics.

Our Vision

We see AGF as complementing existing EA work by focusing on practical implementation. Rather than duplicating research efforts, we aim to provide technical capacity—data engineering, ML development, and deployment support—that helps translate EA insights into working systems for governments and nonprofits.

This is our theory of change: If we can empower the public sector—especially in low-resource environments—with intelligent, transparent, ethical AI tools, we can massively improve institutional decision-making, reduce inequality, and ensure that the benefits of AI are distributed justly.

If that interests you, we want to talk. You can reach us via our website or email us at info@algorithmicgovernance.org.