Article Note: On Libraries Using Alexa and other similar devices

This article reports on a study of libraries that use voice assistant technologies somehow. Voice assistant technologies are things like Amazon Echo, Google Home, etc. Library uses can include in their services, programming, and even for check out procedures. The study aims to see where libraries are using these technologies, to assess whether library workers are aware of privacy concerns related to these technologies, and to see if those workers are doing anything to educate their patrons not just on using the devices but on privacy issues. 

The authors sent out an online survey in Fall 2019 to 1,929 public and academic libraries via e-mail. The survey instrument is included with the article. How many did they get back? 86 total responses. Talk about a seriously small sample, although this kind of low numbers are not uncommon in LIS articles. Most of the respondents (61%) were from public libraries. The rest were from academic libraries (38%) and one lonely school library (1%). Oh, and by the way, out of those 86, only five respondents (6%) "reported that their library is currently using Amazon Echo, Google Home, or Apple Siri devices for patron services and programming" (5). This is barely a blip on the radar so to speak for something the American Library Association reported as a "trend" in a 2017 report they put out. Overall this article mainly looks at what a very small number of libraries are doing.

 

As I was about to read this, I was reminded that a family member got us the Google device as a holidays gift some time back. Given I already know how invasive and, let's be honest, creepy even these devices can be the gift stayed in its box unused, and it will stay there most likely. Before reading this article, I had no idea these would be deployed in libraries, but I am not really surprised. There are plenty of librarians out there who'll try anything "edgy" or cool in tech for the coolness factor, and even to get an article or two out of it (like this article I am reading now). 

Some highlights from the article with my comments: 

• "Nicole Hennig points out there are tremendous opportunities for voice assistants to assist 'people with disabilities, the elderly, and people who can't easily type'" (1). 
• I can see this point, but I wonder if it really justifies ignoring privacy concerns. Let's be honest, there are no guarantees of privacy protection. 
• "There are many creative and potentially helpful ways that voice assistants could be integrated into the library setting, including enhancing read-along with music and effects, providing accessible services for elderly patrons or individuals with disabilities and providing an alternative access point for common library queries and institutional information (e.g. searching titles, placing holds, requesting library event information)" (3). 
• "At the same time, the key privacy issues inherent in voice assistants are often downplayed as secondary concerns while librarians are encouraged to press forward and experiment with smart technology adoption" (1).
• See my remark above. This is oh so 2.0 (or twopointopian to borrow from the old Annoyed Librarian back in the day). Some things really do not change. 
• Definition of a voice assistant: "Voice assistants are a type of digital assistant technology, also known as virtual assistants, and can be broadly defined as computer programs designed with human characteristics that act on behalf of users in digital environments using voice interfaces. Apple's Siri, Microsoft's Cortana, and Amazon's Alexa are prevalent examples. . . " (2). 
• Keep in mind that these technologies do record and capture a lot of private and intimate user information including, but not limited to "biometric data (e.g. voice recognition), consumer habits, internet-based transactions, personally identifiable information (PII), and geographical information" (3). 
• You should also probably be concerned over the fact they can be "always on." 
• You may also want to be concerned how law enforcement could get access to these record and how they would use them. 
• Oh, and you may also want to be concerned not just about the big companies but the contractors those companies often outsource part of their work to. 
  
• But hey, there are so many creative and helpful ways to use these technologies in libraries. 
• Speaking of Amazon: "Lastly, Amazon is a known collaborator with US government agencies like Homeland Security and Immigration and Customs Enforcement (ICE), hosting their biometric data on Amazon Web Services (AWS). Amazon has a reputation for being one of the least transparent technology companies in terms of data sharing practices, and has routinely evaded questions about if/how much of customers' Echo data has been turned over to federal authorities" (3). 
• So much for some librarians as safeguards for privacy. Just invite Amazon right in. 
• Another use mentioned in the article: for technology "petting zoos." Keep in mind this was five libraries. "When it comes to using these devices in library programming, the most common response was for use in technology petting zoos and in technology classes where patrons can see technology demonstrations and ask library staff questions, or get one-on-one tutoring sessions. . ." (6). 
• So it seems much of the use is in education, as in showing patrons how to use the new toy they got over Christmas/holidays so they can use it at home. There is precedent here as patrons will often bring things like new Kindles to libraries expecting them to show the patrons how to use them and fill them with books. 
• "The two libraries who reported circulating voice assistants indicated that they did not include any privacy information with voice assistant devices at checkout" (7). 
• Topics for data literacy classes: ". . . understanding your personal risk profile; password managers and security; how to understand and protect your digital footprint; and sessions on Facebook and Google where staff  'walk uses through how to find their information and make decisions about it.' Several respondents identified information literacy topics in conjunction with data literacy, noting that their library teaches classes about identifying 'fake news,' phishing scams, and evaluating the authority of websites and website content. None of the responses specifically named issues around privacy or data capture by voice assistants or other smart technologies as topics covered in library technology clases" (7). 
• Someone should teach this kind of thing in libraries. It certainly should be taught consistently to high school and college students, including here. I did a small talk on data literacy to a group of students in a campus retreat some years back, and that was a one time thing; I was never invited to later iterations of the retreat, the topic was basically dropped. In my humble opinion, this does need to be taught, and it needs to be more than just putting some information on the college website and hope students will find it somehow, let alone read it. We can add some of this to our current information literacy curriculum, but it also needs to be a coordinated teaching effort with stakeholders like faculty and campus IT. 
• Some library workers commented on issues of labor for things like having to clear private data and "scrub" clean devices they lend. This is applicable to those checking out voice assistants as well. "This comment points to the extra staff labor that underpins technology services, which is often not considered part of infrastructure for offering these services" (10). 
• Possible legal issues library workers have identified: possible Family and Educational Rights and Privacy Act (FERPA), library liability, and "librarians' ability to uphold first amendment rights with voice assistants" (10). 
• "The assertion that library core values may be in conflict with the technology providers that are designing voice assistants is very astute, and important for libraries to consider when weighing the decision to experiment with these (and other) emerging smart technologies" (11). 
• That was a statement of the obvious. 
• "Our research suggests that library use of voice assistants poses many-as-of-yet unresolved privacy issues for library staff and patrons alike" (10). 
• No shit, Sherlock. 
• Key considerations the article identifies for library professionals: 
• Circulation procedures. 
• Third-party digital content platforms. This is how those devices integrate with some of libraries' vendor systems, like say Overdrive. 
• Education in digital literacy. By the way, ALA defines digital literacy as: "the ability to use information and communication technologies to find, evaluate, create, and communicate information, requiring both cognitive and technical skills" (qtd. in 12). 
• Do note: "However, this definition of digital literacy falls short of considering the role of assessing data collection, storage, and use as a core part of digital knowledge. Expanding digital literacy training, both for staff and patrons, to include awareness of data ecosystems and privacy concerns that undergird smart technologies is a must for responsive library services" (12-13). 
• "Given the incredibly low bar for federal access to information under the USA PATRIOT Act, actively facilitating the collection of patron and staff interactions, particularly without informed consent, should give librarians pause" (13). 
• I wish the above was just alarmist stuff, but it is not. It is the times we live in. 
• "As it stands, these technologies are not currently filling a gap in library services that cannot be otherwise met by more traditional service models that carry fewer potential harms for our patron communities" (13). 
• Let's be blunt here. For the "ooh so creative" and "helpful" things these things may or not do, they are just not worth the risks and security issues. In other words, other than being flash, this technology is something we can and should do without just fine. 
• In fact, the authors' conclusion agrees with me: "Indeed, we encourage library workers to consider informed non-adoption of these technologies as a socially responsible professional stance until key issues we have outlined are addressed" (14). Since I do not foresee those key issues being addressed by the government nor the companies running those technologies any time soon, I would say it pretty much means non-adoption period.
  
• And another reason to do without: "Libraries need to consider that these tools, as extensions of policing data networks, may directly endanger, particularly, Black, LatinX, and indigenous people who are already subjected to over policing" (13).