The digital turn is also impacting on the way research is conducted, including its quality assurance. AI systems offer new possibilities for information processing and text support which, when used responsibly, can help ease workloads. For this reason, the DFG has decided to permit the use of AI systems in the review process – subject to clearly defined conditions. The underlying principles here are confidentiality, transparency, quality assurance and responsibility. These are set out in detail in a binding guideline on the use of AI in the review process.

The use of AI systems must be disclosed in a manner appropriate to academic standards. “Disclosure” currently means specifying which generative models were used, for what purpose and to what extent, for example in summarising the state of research, developing a research method, analysing data or generating hypotheses. AI tools that do not affect the content of the proposal (e.g. grammar, style or spell checks, translation tools) need not be disclosed. Specific marking of the passages in question in the proposal is not required.

All reviewers, members of statutory bodies and of course the DFG Head Office are obliged to comply with the principles set out in the Guideline on the Use of AI in the Review Process(interner Link). This includes the obligation to ensure that the content of a proposal does not enter the training data of the AI system being used or is put to use by the provider of the AI system or third parties for other purposes. Data protection requirements must also be observed.

If you are not acting as a reviewer or as a member of a statutory DFG body, or if you do not use AI systems in these processes, accepting the conditions has no impact on you. Please note that these rules may be relevant in the future if your role changes, however. If you are asked by the DFG at a later stage to review a proposal, for example, you will already have given your consent to the conditions governing the use of AI systems in the review, evaluation and decision-making process when registering in elan.