The legal and ethical issues affecting the acquisition or creation, processing, preservation, and access of digital content are important context to include in digital curation training. They provide a basis for decision-making and workflow development.
Accessibility means that people with disabilities can still access our collections, although improvements in accessibility typically improve the user experience for everyone. For example, we might evaluate how well our systems work with screen readers, if color schemes are suitable for someone with color blindness, or how the amount of content would work for someone susceptible to information overload. While this is most often done with access systems and users in mind, it also applies to how back-end systems work for staff with disabilities. If users must come on-site to access digital content, we also need to address potential barriers to navigating the building.
Cultural sensitivity means being aware of and respecting differences between cultures, including the knowledge practices of the originating culture of our collections. Knowledge practices include things like how information is organized and who may have access to it. It is most commonly applied to indigenous records, through appropriate access restrictions, descriptive terminology, and knowledge labels. These principles can also apply to working with local communities that have distrust of UGA.
How we describe our collections, including digital collections, influences how they are discovered and understood. There has been a lot of work in archives on reparative description in the last few years, seeking to address harm and bias in existing description that was often created in the name of neutrality. We don’t, for example, want to make someone search for a common but racist term for their own community to find collections related to themself. A lot of focus is given to race, gender, and sexuality, but this can apply to any terms that describe people. As much as possible, we want to describe communities in terms they would use to describe themselves.
Preservation activities like storing many copies and regularly running processor-intensive tasks like fixity verification require large amounts of energy and rare materials to produce the hardware. In practice, this can look like adding energy efficiency as a requirement when making new hardware purchases, revisiting the frequency and method of tasks like fixity verification, and using appraisal to limit the amount of content we commit to preserving.
Digital collections, especially those indexed by search engines, increase the chance of a user unexpectedly coming across images or language that they find harmful or offensive. They may see snippets in search results or images in thumbnails before they have a chance to see the context and become aware of the likelihood of finding something harmful. Mitigation options include harmful language statements, caution in assigning thumbnails, and blurring images unless someone chooses to click on it. The goal is not to restrict access, but to give users information about what they are about to encounter so they can choose if they want to engage with it or not.
This kind of content can also affect employee mental health. We might give someone the option to not work on a particular collection, to take breaks, and make sure they are aware of mental health resources for UGA employees.
Once the Libraries is responsible for the content, we need to ensure that it is kept secure. All storage should be protected from hacking, accidental unauthorized access, and accidental or malicious deletion. The physical locations of storage hardware should be locked and have appropriate HVAC to limit the risk from fire and water damage.
Copyright limits what we can do with digital objects, including making copies in multiple storage locations or different formats for preservations and providing access. Deeds of gift, licenses, or other legal instruments need to transfer sufficient rights to the Libraries for us to be able to do our jobs.
Privacy is of greater concern with digitized and digital archives because sensitive and personal content is easier to discover once it is full-text searchable, especially if it is indexed by search engines. This ranges from social security numbers and other PII that might enable identity theft to legally protected medical and student information to more nebulous private facts that someone would not want released.
In addition to the creator and/or donor of the records, privacy also applies to the individuals who are the subjects of our collections, especially for those who had no control over the records creation or their donation the archives, and likely are not aware that there are records about them in our holdings. This disproportionately affects minority groups.
The likely presence of this kind of information influences whether something is a candidate for digitization, the amount of review required for digital archives during processing, and the method for providing access. We can reduce these risks by discussing likely private information with donors, using digital forensics and other automated methods to support (but not replace) staff review for sensitive content, and publishing clear access and take down policies.