Researchers, journal editors and scientific institutions should work together to improve communication about misconduct cases. Although published retractions are logged by PubMed and other databases, and by blogs such as Retraction Watch (http://retractionwatch.wordpress.com), the scientific community needs a way to identify flawed articles that have not been formally retracted but have been assessed as containing falsified data or having ethical problems (see, for example, Nature 476, 263–264; 2011). To this end, we have piloted an open database of publications for which misconduct has been established by committees (such as offices of research integrity within research institutions). The database is collaborative and is coupled to an online platform on which scientific integrity can be openly and constructively debated (see http://www.scientificredcards.org and T. Flutre et al. Eur. Sci. Ed. 36, 51–52; 2010). The website focuses on the publications and not the authors, to avoid 'naming and shaming'. It has been legally validated by the French National Commission on Informatics and Liberty, so that such information can be made public while respecting privacy laws. To expand this initiative, the legal implications would have to be considered. It would need to be endorsed by the research community, which would cooperate to maintain and moderate it. Extensive publicity would be essential to ensure that the facility is used effectively. Our pilot project offers a route to reinforcing society's trust in science. Creating a public library of misconduct through a collaborative web platform is a timely, transparent and efficient way for the research community to communicate about possible scientific impropriety.