In addition to any data management or sharing required by funding agencies, all data producers and users must be aware of and abide by ethical and legal guidelines.
Data is governed by copyright law. See the Libraries' Copyright guide for extensive information and links. Although raw data or "facts" are not copyrightable, an arrangement of data within a database, or a selection or expression of data (e.g., in a table) may be copyrighted.
Here is an interesting discussion of these issues in the context of government data.
In response to the White House's Directive from the Office of Science and Technology Policy to expand public access to the results of federally funded research, federal agences have begun releasing plans that outline the requirements for research funded by these agencies to be made public. These plans apply to both the publications and the scientific data used in the research.
This page from Oregon State links to each agency's Federal Public Access Plan, and is being updated as new ones are released in 2015.
The National Science Foundation now requires that a two-page Data Management Plan be submitted with every grant application. While no across-the-board requirements have been set, researchers are expected to share their primary research data in a reasonably timely and efficient manner. The data management plan should facilitate such sharing, and the plan will be considered as part of the overall merit of the grant. Presumably, reuse and innovative applications of the data would show broader impact of the research, one of the criteria by which grants are evaluated.
The NSF states that:
"This supplement [data management plan] should describe how the proposal will conform to NSF policy on the dissemination and sharing of research results (see AAG Chapter VI.D.4), and may include:
Each directorate of the NSF is drafting its own data management guidelines, which are linked here. The norms of each discipline will guide what standards and sharing mechanisms will be expected for NSF-funded data. The NSF FAQ page provides further guidance.
See also, the Association for Research Libraries' Guide to the NSF Data Sharing Policy.
Other agencies also require that research data produced by a grant-funded project be made publicly available. Notable examples are:
Further information on funding mandates, along with an excellent bibliography, can be found at the University of Minnesota's Funding Agency and Data Management Guidelines page.
Here is another summary of data requirements from various funders.
Data, especially involving human subjects, should be collected in an ethical manner, stored securely, and closely reviewed before distribution to avoid the disclosure of confidential information.
Rutgers researchers should comply with the ORSP guidelines for Human Subjects Research.
Health research is subject to HIPAA rules.
ICPSR has information on maintaining confidentiality and checking for human subject identifiers when evaluating a public release version of data, and how to distribute sensitive data under restricted use contracts.
The UK Data Archive's Data Security page contains useful guidance on storing and protecting confidential data.