SUBSCRIBE NOW!

Vol. 43 No. 2 — March 2023

THE SYSTEMS LIBRARIAN
Libraries Are Not Neutral—Neither Is Technology
by Marshall Breeding

Values Shape Library Services

Broadly speaking, libraries perform the vital role of delivering information resources to their communities effectively and efficiently. There are differences depending on the type of library, each with its own challenges and constraints. Libraries in higher education build collections to serve the teaching and research of their specific institutions. K–12 school libraries provide resources for the classroom curriculum and for general literacy and interests of students. Public libraries address the incredible challenge of delivering resources of interest to their entire communities, spanning all ages, cultures, and identities.

Libraries strive to provide resources for all segments of their service populations. Building library collections relies on processes informed by data-driven decision making. Library workers endeavor to meet the information needs of their constituents to make decisions to acquire, retain, or deselect items that comprise their collections. This work might be based on the personal knowledge and expertise of librarians, and it may make use of technology-based tools and services. Given limited resources, selections must be carefully prioritized. Tools for collection analysis help libraries to build collections with the strongest impact and to identify gaps or community segments that may be underserved. Professional judgment is an important ingredient in the process of creating and providing access to library collections. Core principles include facilitating access to resources vetted by domain experts and actively avoiding misinformation.

Libraries likewise disassociate themselves from hate groups or other entities that incite harm or violence. A neutral position would grant access to all individuals regardless. Some circumstances may warrant denying services to those who promote misinformation, harm, or violence. Libraries work to be inclusive, providing materials to all segments of the populations they serve. They resist censorship. In recent years, libraries in the U.S. have seen new waves of attempts to restrict or ban materials from library collections. The efforts of libraries to resist censorship can be stymied when the government organizations that fund them or boards that oversee them place restrictions into legislation or official policies. Attempts by special interest groups to restrict material available to all readers runs counter to the core values of libraries but, in some cases, have become realized.

Privacy is a fundamental value for libraries. They follow procedures that fully protect the privacy of their users, including any personal details and especially information relating to the resources and services accessed.

Values Are Embedded in Technology

Technology is not neutral. Rather, technology-based products and services operate according to a design or coding that reflects pragmatic values. It’s essential for libraries to understand the performance and internal workings of the technologies they implement and ensure consistency with professional values and institutional policies.

When selecting and implementing technologies, libraries should evaluate how they operate relative to their distinctive policies and values. Libraries do not seek to monetize their services. Rather, they follow a nonprofit model to deliver information resources and related services without cost to their users. Exceptions apply, naturally, but providing free access to materials is what sets libraries apart from bookstores and commercial content publishers or distributors.

The Neutral Zone

Some layers of technology can be considered neutral relative to concerns for ethics and values. Lower-level infrastructure can generally deliver data and content without introducing problematic characteristics or behaviors. ISPs provide raw bandwidth to users and manage the flow of data throughout the internet. The routers and other equipment involved at this level deliver data efficiently and generally without bias. The practice of Net Neutrality remains in place, so that the content of all organizations is provided equitably. The choice of an ISP is based on factors such as cost, bandwidth levels, and reliability, mostly apart from ethical concerns.

Web standards, such as HTML and CSS, deliver and present content and are robust specifications that are designed to work well for any type of organization. Correct implementation of the current versions of these standards will ensure that content can be consumed consistently on laptop or desktop computers, tablets, or smartphones and across all major OSs. All pages should be validated for syntactical errors, for mobile-friendliness, and for accessibility. Validity at this level does not guarantee accessibility, which also depends on design and implementation decisions.

HTTPS manages the flow of content across the internet and ensures that it is encrypted and transmitted securely. The unencrypted version, HTTP, should not be used since it allows for content to be intercepted and captured as it moves across networks and is inconsistent with library values and policies regarding patron privacy. When using HTTP, unknown third parties can access information as patrons use a library website, including search terms, results listed, and items accessed.

Higher-Level Concerns

Technologies that operate on top of the core internet infrastructure level require much more careful consideration related to privacy and other policies of special interest to libraries. CMSs (such as WordPress and Drupal) enable organizations to build websites without working directly with HTML, CSS, JavaScript, or other coding conventions. These environments do not necessarily introduce concerns directly, but may be configured to invoke components that interact with advertising networks or other entities that can jeopardize the privacy of those accessing the site. Libraries should carefully review any content or code that these tools may layer onto each page, including tags for analytics and JavaScript that may come with some unintended actions.

Ad Networks

The economy of the web depends, to a large extent, on advertising. Much of the content we consume on the web includes ads, usually selected according to interests or characteristics of the specific person viewing the page. A layer of infrastructure operating behind the scenes supports personalized placement of ads and the flow of data and money between advertisers that want to place ads and the companies that sell ads. These advertising networks comprise components that collect data about users, massive data warehouses, data brokers that buy and sell personal information about users, and systems that perform real-time auctions to place ads on a page. This advertising infrastructure pervades the commercial web and beyond.

Ad trackers can pass sensitive information from a library user into the advertising network. This information may not only expose personal details about the user, but can also include the search queries or items viewed. Libraries interested in protecting the privacy of their users need to take extraordinary care to ensure clear separation from their services and the advertising infrastructure. These measures mostly involve avoiding the insertion of the trackers and other code snippets that cross the boundaries between a library’s own website and information services and those of the advertising networks. These trackers can be inserted through CMSs—often inadvertently—or via the inclusion of JavaScript code that may not be well-understood. Libraries should check their sites for the use of third-party cookies, digital fingerprinting, and other techniques that can expose details about their users.

Few libraries include advertising trackers intentionally, but they can be difficult to avoid without using tools to detect them. Some helpful tools for detecting ad trackers include Ghostery (ghostery.com), a browser plugin that detects and disables unwanted trackers, and Blacklight from The Markup (themarkup.org/blacklight), which is a service that can scan any website and report on characteristics such as trackers, third-party cookies, or other techniques that invade privacy.

Analytics

Organizations such as libraries need to understand use levels of their web-based resources, and to do this, they usually employ services such as Google Analytics. Data I have collected indicate that almost all public and academic libraries use Google Analytics. Deploying this tool is a gray area regarding library privacy practices. Doing so involves enabling a tag that transmits data to Google each time a user views a page, including data about the user and the resource accessed. The intrusiveness of Google Analytics can be mitigated by enabling a setting that anonymizes the IP address of the visitor.

Social Networks

How libraries use social media on their websites and other resources brings up a variety of issues. It has become a widespread practice for libraries to include links or icons inviting their website visitors to connect to their social media sites. From a privacy perspective, these links can be problematic since they can easily carry personal information about the patron and their activity on the library site to the social network. The referral URL transmitted to the network often includes search strings performed on the library site. The presence of the Meta Pixel can notify Facebook about each page accessed on the library website, including those that involve search and retrieval of information resources. If the user also happens to be logged into Facebook or Instagram, even passively, very specific information about that user is associated with interactions on the library website.

Outbound linking from library resources to social networks also involves encouraging visitors to the library website—hopefully, carefully constructed to conform to strict privacy practices—to exit into an entirely different type of environment that thrives on the extraction of personal data and follows an ad-based business model. This scenario differs from links to information services and content resources, which are vetted for both trustworthy content and for privacy policies consistent with libraries. The privacy practices of social networks contradict the terms of most library privacy policies, but the presence of social media links and icons has become ubiquitous on the websites of U.S. public and academic libraries.

The hype of the Web 2.0 era led many libraries to fully embrace social media. The basic idea involved developing a presence among the virtual gathering places where their communities congregate. Having an active presence on social media sites that promote the library and lead individuals to library resources continues to be a sound strategy. But to invite those who have already landed on a library site to leave and step into more treacherous ground seems counterproductive in many ways.

AI

AI has made tremendous advances in recent years. In many contexts, AI-based tools can provide benefits in automation as well as in information services. ChatGPT, for example, shows extraordinary capabilities for generating text that is hard to distinguish from that written by humans in many disciplines and topics. Such AI services are naturally driven by the content on which they are trained and may also introduce elements of bias. Although they are powerful, a library should examine AI-based technologies carefully to ensure they operate within the bounds of objectivity and transparency that would be expected in any service it provides.

Blockchain

Blockchain has also attracted massive attention, both as a tool for managing transactions and as the basis for cryptocurrencies. The foundation of blockchain involves open ledgers that use encryption technologies to lock in any transaction so that it can never be changed. Blockchain implementations require substantial computing resources, which can rise to a massive scale for implementations such as Bitcoin that are based on mining as a proof-of-work in the generation cryptocurrency value. The key issue for blockchain in a library context relates to the openness of transactions, which contrasts with the privacy required for those involving access to library resources. Although it is possible to add layers that anonymize transactions on a blockchain, it is hard to see that the technology is well-suited for most library applications. The computing resources required and the associated environmental impact of the energy consumed also raise concerns.

Nonprofits in a Commercial World