Open data and open government are great initiatives that are increasingly being adopted by Canadian municipalities. Many municipalities and provincial governments license their open data under a modified variation of the Open Government License (UK 3.0).
The license itself is generally fairly permissive in practice. It allows you to copy, publish, transmit, adapt and commercially exploit OGL-licensed information in return for an attribution statement.
However, in recent times, this license has been complicated through freedom of information and privacy laws. One principle aspect of freedom of information is that it is not absolute.
Freedom of information and privacy laws dictate what information that governments may legally withhold, must withhold, or must withhold until consulting with relevant parties. This can range from contractual obligations, to sensitive information, to personal and private information.
In many ways, this is a bit confusing. If governments deliberately go out of their way to release open data, why would open datasets include information that could or must be withheld? Would it not be reasonable to conclude that the information released has nothing that is subject to an exception?
The issue lies in the details. How can you be sure that a OGL-licensed dataset with freedom of information provisions is truly free of exempt material?
Government employees are not infallible and make the same mistakes as you and I.
Therefore, OGL licenses with provisions for freedom of information laws are non-free licenses, that present limitations on its full and absolute use. If you create a product that inadvertently contains information that cannot be released under freedom of information laws, you put your product in a legal grey zone, because it could contain data that you were not legally licensed to use.
One potential workaround, is to make freedom of information requests for already-released open datasets. This effectively ensures that each released dataset is properly reviewed by a freedom of information specialist. The response to your request would confirm that the data can be used in full because it contains no information subject to exceptions, or you would receive a sanitized dataset with all exceptioned data removed.
However, I’m not a lawyer and I can’t say with any certainty whether this could be a final solution that absolves users of modified-OGL licensed content from blame and liability