AI in the Review Process
The DFG permits the use of artificial intelligence (AI) in the preparation of reviews. However, such use is only permitted subject to clearly defined conditions, which are set out in a guideline. Reviewers may use AI systems for support purposes at individual stages of their work, for example in structuring their own notes, editing language and screening publicly available literature. However, the academic analysis, evaluation and recommendation remain the responsibility of the reviewer.
- Guideline on the Use of A(externer Link)
- White Paper: AI in the Review Process – DFG Position and Perspective(externer Link)
- Contac(externer Link)
What’s New
On 22 April 2026, from 10:30 to 12:00, the DFG will host an online information event on the topic of “AI in the review process”. The login data will be provided here.
The DFG regards the AI systems listed below as being currently compatible with the guideline (as of March 2026). The aim of this approved list is to provide guidance for reviewers. The list will not be updated on an ongoing basis.
- Locally hosted AI systems, i.e. AI systems installed on the reviewer’s own device where processing takes place exclusively on that device, without cloud computing or the transfer of processing results to the cloud. Example: https://ollama.org(externer Link)
- AI systems hosted by a trustworthy institution, e.g. a university or a public research institution. Example: http://kiconnect.nrw(externer Link)
- Cloud-based (commercial) systems subject to a contractual licence agreement, i.e. AI systems for which a contractual licence agreement is in place guaranteeing data security, excluding use of the input data by the provider (e.g. for training purposes) and ensuring that these conditions are actively set by means of the appropriate security settings (e.g. “no logging”; “no storage”).
Guideline
The DFG has adopted a guideline which sets down four principles governing the use of AI in the review process. The guideline is available here:
Principles
See here for a brief overview of the purposes for which AI systems may be used and the four principles set out in the guideline:
Human analysis and the expert assessment of the proposal are core elements of the review process. The use of AI systems in the preparation of reviews is therefore permitted only in a supporting capacity. All substantive judgements regarding the funding recommendation rest solely with the reviewer.
AI systems may be used only where care is taken to ensure that confidential content of funding proposals is not processed without oversight, stored, or used for purposes other than the conduct of the review process. The supportive use of AI in drafting a review must not result in the disclosure of confidential proposal content. This can be ensured by using the following types of system:
- Locally hosted systems without cloud computing
- Systems hosted by a trustworthy institution (e.g. a university or public research institution)
- Cloud-based (commercial) systems subject to a contractual licence agreement guaranteeing data security
If an AI system is used in preparing the review, this must be disclosed, specifying the purposes of use.
Uncritical adoption of AI-generated text is not permitted. AI-generated text must always be critically assessed, adapted and checked for factual accuracy, currency and potential bias structures.
Full responsibility for the content of the review lies with the person who prepared it. This applies even where parts of the text were generated with the assistance of AI systems. Responsible expert review requires subject expertise and ethical judgement, which by their very nature AI systems do not possess. Responsibility must therefore not be transferred to an AI system; it remains with the reviewer.
White paper
The DFG has issued a white pape(externer Link) summarising its internal process of clarification regarding the use of AI systems in the review process.
Contact
| E-mail: | KI@dfg.de |