The Impact AI Method assesses the impacts of AI use cases that pursue "AI for good" goals, for instance by contributing to sustainability or the public interest. Its holistic yet practice-oriented approach examines AI systems already in operation and asks which implications they have for people, society and the planet.The method was developed within Impact AI, a transdisciplinary research project at the Alexander von Humboldt Institute for Internet and Society (HIIG) in Berlin, in cooperation with Greenpeace Germany and Economy for the Common Good (ECOnGOOD). Over a five-year period, the research team led by Dr. Theresa Züger is applying the framework to AI use cases that are independent of the project.This record brings together the full set of materials required to carry out an assessment. The overview document outlines the seven-step process and shows how the individual instruments fit together. Each instrument is also available as a separate record:Overview Document – the seven assessment steps and how the materials connect
Interview Guides (Organizational Interviews) – a pre-questionnaire plus three guides on governance and impact, sustainability, and risk and mitigation (10.5281/zenodo.23187591)
Technical Survey/AIMS-Templates – templates documenting the AI system, its models and datasets (10.5281/zenodo.23188065)
Technical Interview Guide – follow-up interview with the technical lead (10.5281/zenodo.23188331)
External Impact Interview Guide – sample questions for users and other external stakeholders (10.5281/zenodo.23188633)
Catalogue of Assessment Criteria and Questions – the criteria, based on the UN Principles for the Ethical Use of AI and extended by public interest and sustainability aspects, used to rate the findings (10.5281/zenodo.23188992)Answers to common questions – for instance on the assessment criteria, the rating scale, the reports produced and how AI projects can take part – are collected in the Impact AI Method FAQs.
