Because of the difficulty in deciphering the contours of copyright protection in scientific data and databases, it is very difficult for the data provider and the user to know when the license is valid and when it does not apply. In other words, it is difficult to know when the imputation is mandatory. There are a number of risks involved. The most fundamental form of imputation is the indication of the identity of the copyright holder, often in the form of copyright © [year] [name of copyright owner]. Maintaining such a notice was an immutable requirement to prevent a work from being made public; this changed in the United States on March 1, 1989, when the copyright registration and signage requirement ended. Copyright-owner-attribution is not necessary in most countries of the world, because of the Berne Convention. On the one hand, there is a risk that data providers will be misled about what they get when they license their data. They may believe that when they apply a license to their data, any use of the data requires an assignment. As I have already explained, this is not the case. If the data is in the public domain or if the use of copyrighted data is a matter of fair dealing, the duty to allocate is not triggered. The attribution is in copyright an acknowledgement of the copyright holder or the author of a work. When a work is protected by copyright, there is a long tradition of the author requiring an imputation while directly quoting parts of works created by that author.
In addition to legal uncertainty, licenses also create a risk of imposing heavy allocation requirements. Especially in the scientific context, projects often rely on data from a wide variety of sources. Depending on the license used, it is possible that each individual or institution that contributed to the project data will be affected. This is a problem we call stacking attribution. In other words, a work in the public domain that is not protected by copyright is not subject to such an imputation obligation in most parts of the world. This is the distinguishing factor between plagiarism that is not crimes and copyright violations, which may be a ground for legal action by the perpetrator. Like licences, contracts suffer from a number of potential disadvantages. On the one hand, they give confusing obligations to users who receive data from a large number of sources, all of which are subject to different usage agreements. This problem is even more pronounced in the case of contracts, at least public licences being somewhat standardized. These are not user agreements, which means that each data source probably has a different usage agreement, filled with legal endowments and other obligations for users.