Identifying and Countering Fake News: New Study Published

Fake news has become a controversial topic, with media organizations, scholars, and even the President of the United States debating the issue. However, it’s not clear what counts as “fake news.” This makes it difficult to diagnose the social harms from fake news, or to create solutions to them. A new report, “Identifying and Countering Fake News,” brings much-needed rigor and clarity to the problem. It is authored by three media and Internet scholars at the University of Arizona: Mark Verstraete, Derek E. Bambauer, and Jane R. Bambauer. The report identifies several distinct types of fake news, including hoaxes, propaganda, trolling, and satire. The study also proposes a set of model solutions to reduce the production and dissemination of fake news.

In the public discourse, “fake news” is often used to refer to several different phenomena. The lack of clarity around what exactly fake news is makes understanding the social harms that it creates and crafting solutions to these harms difficult. This report adds clarity to these discussions by identifying several distinct types of fake news: hoax, propaganda, trolling, and satire. In classifying these different types of fake news, it identifies distinct features of each type of fake news that can be targeted by regulation to shift their production and dissemination.

This report introduces a visual matrix to organize different types of fake news and show the ways in which they are related and distinct. The two defining features of different types of fake news are 1) whether the author intends to deceive readers and 2) whether the motivation for creating fake news is financial. These distinctions are a useful first step towards crafting solutions that can target the pernicious forms of fake news (hoaxes and propaganda) without chilling the production of socially valuable satire.

The report emphasizes that rigid distinctions between types of fake news may be unworkable. Many authors produce fake news stories while holding different intentions and motivations simultaneously. This creates definitional grey areas. For instance, a fake news author can create a story as a response to both financial and political motives. Given this, an instance of fake news may exist somewhere between hoax and propaganda, embodying characteristics of both.

The report identifies several possible solutions based on changes to law, markets, code, and norms. Each has advantages and disadvantages. Legal solutions to fake news are likely to conflict with strong constitutional (First Amendment) and statutory (section 230 of the Communications Decency Act) protections for speech.  Market-based solutions are likely to only reach a subset of fake news. Code solutions may be limited by the difficult judgments required to distinguish satire from other types of fake news. Norms and other community solutions hold promise but are difficult to create through political will.

Some types of fake news are more responsive to regulation than others. Hoaxes are produced primarily in response to financial motivations, so solutions that remove (or decrease) the profit from fake news stories are likely to reduce the number of hoaxes created. By contrast, propaganda is produced primarily for non-financial motivations, so changes in its profitability are unlikely to significantly reduce its output.

The report introduces several solutions that can serve as starting points for discussion about the practical management of fake news, and networked public discourse more generally. These starting points include: expanding legal protections for Internet platforms to encourage them to pursue editorial functions; creating new platforms that do not rely on online advertising; encouraging existing platforms to experiment with technical solutions to identify and flag fake news; and encouraging platforms to use their own powerful voices to criticize inaccurate information.

For more information, contact lead author Mark Verstraete at markverstraete AT email.arizona.edu.