Call for Applications: Impact AI Fellowship 2027

Das Alexander von Humboldt Institut für Internet und Gesellschaft (HIIG) öffnet 2027 seine Türen für neue Impact AI Fellows! Indem es eine neu entwickelte, innovative Methode zur Wirkungsmessung in die Praxis umsetzt, bietet das Fellowship-Programm den Teilnehmenden die einzigartige Gelegenheit, die tatsächlichen Auswirkungen von KI-Anwendungen auf Gesellschaft und Umwelt zu untersuchen. Interdisziplinäre Forschung ermöglicht es den Fellows, Evidenz zu erarbeiten und zur Debatte über KI für Nachhaltigkeit und Gemeinwohl beizutragen. Während des zehnwöchigen Fellowships vom 16. August bis 22. Oktober 2027 evaluieren die Fellows eine konkrete KI-Anwendung.

About Impact AI

The transdisciplinary research project Impact AI, located at HIIG, has developed the Impact AI Method, a holistic and structured impact assessment. The project works in partnership with Greenpeace Germany and Economy for the Common Good (ECOnGOOD). The method offers a holistic and structured impact assessment. By applying it to individual AI projects, we generate evidence on the societal and environmental impacts of using AI systems and bring new insights to the debate on AI for sustainability and the public interest.

Learn more about our Impact AI Method and previous impact assessments in our Impact AI FAQ.

What fellows can expect

The aim of the fellowship is for participants to learn the Impact AI Method and apply it in practice. To do so, they will conduct a complete impact assessment of an AI project, from preparing and conducting interviews through data analysis to writing a final assessment report. The programme combines workshops, expert input, peer exchange and independent work. Throughout the fellowship, fellows will receive close guidance and support from the Impact AI research team. 

In the first week, fellows will be introduced to our impact assessment method and will assess one real-world case of an ‘AI for good’ application and its impacts within the fellowship timeframe.

Important note: Since interview preparation requires sufficient lead time, the preparatory phase begins three months before the official start of the fellowship with a mandatory online session on May 11 2027.  

What we offer

• Comprehensive training in the Impact AI Method
• Workshops led by researchers and practitioners
• Individual supervision by the Impact AI research team
• An interdisciplinary peer-learning environment
• Lead authorship of a case report in support of our research team
• Opportunities for future collaboration and publication based on the research data from this case study
• A fixed fellowship stipend of €5,000 per person for the ten-week period (this amount includes any tax or social security costs that may apply in the respective fellows’ countries of origin). 

In addition, we will financially support costs related to travel and accommodation necessary for participation in the in-person programme, on an individual basis and as far as our capacity allows. Fellows are responsible for any further expenses (such as health insurance, visa, meals etc.). In exceptional cases, we will consider assisting with additional necessary costs that may arise. 

Your commitment

All fellows will agree to a fellowship contract, and need to commit to the specified activities and to producing the outputs. For two specific weeks (weeks 1 and 7) the fellowship is a full-time in-person commitment in Berlin. In the other weeks, part-time and remote fellowship work will be possible, if other obligations require this (exact conditions can be negotiated individually but need to be realistic for the given fellowship objectives). For all fellows, participation in all in-person sessions and the regular online check-ins is mandatory.

The structure of the fellowship

The fellowship will take place from 16 August to 22 October 2027. The in-person programme will be held at the Alexander von Humboldt Institute in Berlin.

Qualifications

We welcome applications from people working in academia, journalism, civil society organisations, public institutions and companies engaged with AI, sustainability, or the public interest. We are looking for people with diverse academic, professional and personal backgrounds who are motivated to explore the societal and environmental impacts of the use of AI. A master’s degree is generally required. Not having one does not automatically disqualify candidates. However, those without experience and adequate training are less likely to be selected. We are looking for applicants who have research experience or professional expertise in AI, sustainability or the public interest, or who are embedded in an organisational or institutional setting that would allow them to continue using the method beyond the fellowship. Candidates who feel that their background offers experience equivalent to a master’s degree and meets these expectations are encouraged to apply.

While an academic track record is not mandatory, fellows need to have a deep interest in research and be open, motivated and capable of learning and applying a complex research method with adequate rigour. They should be interested in interdisciplinary research and in collaborating with other fellows and researchers, but also able to work independently with a high degree of commitment and reliability.

Fellows can propose a specific AI project they would like to assess. However, this is not a requirement. Through a call for cases, we support fellows in selecting and connecting with a suitable AI project in the field.

Application documents

Applicants should complete the online application form down below and submit the following documents (please combine them into a single file):

• CV (max. 4 pages) 
• Letter of motivation including. level of AI expertise, organisational background and potentially organisational motivation (if applicable) (max. 2 pages) 
• If applicable: Description of the proposed AI project (max. 2 pages) 

If you are interested in the fellowship and would like to learn more, we highly recommend joining one of our online information sessions, on 10 November 2026 at 10 am or 17 November 2026 at 3 pm. To participate, please click here (no registration required).