Although there is no one universal definition of “rural,” in general, rural areas are those outside of population centers and their suburbs. In the US, these areas cover approximately 86 percent of US landmass. Although rural county populations have been in a state of consistent decline, they are home to nearly 60 million people, or a bit less than 1 in 5 US residents. That’s more people than lived in the top six US population centers in the 2020 Census.
Rural communities are each unique, but they share many of the same major areas of concern including economic development, connectivity, education, health, agriculture, and climate. Of course, some of these overlap with the major concerns for urban centers. Yet, it is how these issues manifest and shape the impacts of technologies like AI and automation, as well as how technosolutionists attempt to solve social issues with technology, that should be of significant concern.
In short, rural communities matter. And that means they should matter when it comes to the development of policies on artificial intelligence (AI).
COVID-19 Revealed the Depth of the Digital Divide
Before we can even begin to discuss AI policy, however, it’s important to understand the extent of the digital divide that still exists on basic connectivity. For example, four percent of rural US hospitals closed between 2013 and 2020, exacerbating an already flimsy rural healthcare system. According to the Government Accountability Office (GAO), these closings mean patients have had to travel about 20 miles farther for inpatient care. Telehealth services, where patients access providers via phone or internet, have been touted as a solution for the lack of available services. Many telemedicine services are incorporating machine learning and natural language processing for data analysis and virtual consultations.
But telemedicine requires good connectivity. The same report from the GAO found that in 2019 nearly 17 percent of people living in rural areas lack reliable broadband. It’s important that 2019 was right before the start of the ongoing COVID-19 crisis, wherein many jobs and schooling moved to remote. The lack of reliable broadband has hampered the ability of those without access to carry on with their lives as best as possible.
This issue goes beyond healthcare into all areas of life. Organizations like the National Digital Inclusion Alliance have long worked on digital inclusion, creating programming aimed at ensuring that under-resourced communities are able to participate in the digital conversations. The NDIA’s policy suggestions include recommendations that Congress make the Emergency Broadband Benefit Program (EBB) permanent. The Infrastructure Investment Bill and American Jobs Act (IIJA), passed in 2021, provides $65 billion to the states and tribal communities for broadband funding and activities that close the digital divide. But many localities lack data on broadband availability, which may lead to communities being overlooked for funding.
Rural Communities are Using AI in Novel Ways
But rural communities are not devoid of AI and automated systems. Agriculture and environmental science are ripe to take advantage of advances in machine learning (ML) systems. My University of Florida colleagues, for instance, are using ML tools to predict the risk of algal blooms that can kill fish and other wildlife, disrupting entire ecosystems. ML is also used in precision agriculture, or smart farming, aimed at increasing productivity and diagnosing plant diseases.
But precision farming and other Ag-ML applications work via data collection and analysis. Many would think that this data is benign and has no connection to individuals. This is incorrect. The constant data collection needed for precision agriculture allows for the triangulation of data from which inferences about individuals and communities. Further, the majority of the precision systems in use are proprietary, and users are at the mercy of the terms of service of the Ag technology providers for protecting their data privacy, governing data, and data ownership. The power imbalance in this data collection, as with other surveillance systems that feed data to AI, is why an omnibus, federal data protection regulation is necessary.
Of course, agriculture is not the only place in rural communities in which AI and automated systems are being implemented. As in cities, rural law enforcement are acquiring and being provided with military-level equipment, including facial recognition software and drones. Some rural housing authorities are using facial recognition systems to surveil public housing residents. The members of rural, urban, and suburban areas deserve protection from the potential harms of these error-prone surveillance tools.
Rural Communities Deserve a Bigger Seat at the Table
AI-related harms will manifest differently in and for different parts of society. To deal with this, and while the sector remains nascent, one of the hallmarks of the White House Blueprint for an AI Bill of Rights is the recommendation that systems be made and audited with “consultation from diverse communities, stakeholders, and domain experts.” These communities are defined as those identified in the White House’s definition of equity and are referenced in the appendix to the Blueprint – a nonbinding framework that delineates five core principles to guide the design, use, and deployment of automated systems within the US. This definition includes those that have traditionally been underserved, such as communities in rural areas.
Contrary to popular belief, these communities are racially diverse, home to 14 million people who identify as Black, Latino, Asian or Native American, according to the Center for Rural Innovation. These are many of the same individuals, then, that the Blueprint seeks to include in shaping the trajectory of this now seemingly ubiquitous bundle of technologies.
But having a seat at the table requires more than mere mentions in frameworks. It requires the implementation of good policy made through participatory consultation. The Biden Administration has implemented its Rural Partner’s Network aimed at job creation, building infrastructure, and economic sustainability. Rural technology, including AI and other automated systems, should also be a topic of consideration.
But this initiative is limited to only 11 states and territories. To truly understand the implications of AI technology for and in rural spaces, and to make better policy, more rural communities must join the discussion. Citizens in rural communities in the South, the Midwest, the Great Plains, and other rural areas)all have something to say.