Core values
The Netherlands Code of Conduct for Research Integrity (NCCRI 2018/2026) defines 5 corevalues governing academic research. Here we summarize the five principles of researchintegrity in the NCCRI and highlight what each implies for AI use in research.
Honesty. The researcher is truthful, correctly describes the research process, does notfake or distort data, does not fabricate sources, and does not present others’ work as one’sown.
Honesty means that the researcher is honest about their uses of AI, correctly describes the role of AI in the research process and does not use AI to distort or fake data, fabricate sources or plagiarize.
Scrupulousness. The researcher uses scientific methods and aims for precision in thedesign, execution, reporting and dissemination of the research. The researcher ensures thatclaims are substantiated.
Scrupulousness means that scientific rigor, soundness, and validity must be warranted when using AI as part of a research methodology.
Independence. Researchers must be able to perform research with academic freedom, freefrom external influence. When conflicts of interest exist, these are to be transparentlydisclosed. The researcher adopts an open and critical attitude, also to their own personalinterests or ideological opinions, and is accountable for them.
Independence means that researchers reflect and report on possible conflicts of interest when they consider to use AI technologies. This includes conflicts that may arise due to profit motives or external influence (e.g. industry collaborations, user lock-in or politically motivated biases).
Transparency. The researcher makes clear what data and methods were used, how theresults were achieved, what role (external) stakeholders played. The research process mustbe verifiable and reproducible. If parts of the research or data are not openly accessible theresearcher needs to justify why.
Transparency means that researchers disclose which AI applications were used in which way, insofar as this AI use impacts the content of the research and the publications based on it. This includes using AI to produce or analyze data and using AI to produce text or other forms of output.
Note: Academic journals may have their own reporting policies and control mechanisms on AI use in research.
Responsibility. The researcher is embedded in society and is accountable to it. Theresearcher respects the environment and the legitimate interests of people and animalsinvolved in the research.
Responsibility means that the researcher considers societal and sustainability aspects of AI use. Based on this consideration, they make an informed choice regarding whether or not to use particular AI systems.