Last Thursday (October 13, 2016) Google announced it is adding a “fact check” feature to Google News, the news aggregator subsite of Google. News aggregators, for those unfamiliar with the concept, “pull together, and allow [the user] to assemble, news from a variety of sources in one place.” Google News, for example, assembles news articles from over 4,000 sources across the globe and organizes these articles according to topic. News aggregators, Google included, allow individuals to customize their “feed” based upon personal preferences, geographic focus, and interests, the idea being that a person can manipulate settings and set preferences to get a variety of news from one initial location.
So how does an entity like Google fact check the tens of thousands of news stories that flow through its aggregator each day? The short answer is that well…it doesn’t.
Despite the amount of attention the development has gotten in recent days, Google isn’t actually out to single-handedly determine the truth of factual assertions in the online news articles we digest every day. Instead, Google has merely introduced a new fact check tag.
Google News has been making use of tags for some time to label stories as “Opinion,” “In-Depth,” and “Highly Cited”—just to name a few—and the new fact check tag merely exists to label an article as one which serves as a fact check to a popular story. Google makes use of algorithms, a barely used “markup” called Claim Review, and the help of entities like the Duke Reporter’s Lab to determine which fact-checking organizations it can trust. For example, PolitiFact appears to pass muster for a variety of reasons, while the fact-checking feature on Hillary Clinton’s presidential campaign website would not due to reasons of obvious party affiliation and agenda.
Of course, it’s reasonable to view a step like this with a certain level of scrutiny. After all, few companies wield the type of influence Google is able to through not only its preeminent search engine, but also its media empire, cell phone service, and growing fiber-optic service. Google News illustrated the power of its influence in a 2008 incident involving a six year old news article on the 2002 United Airlines bankruptcy. On September 7, 2008, the South Florida Sun-Sentinel erroneously listed a 2002 article about a United Airlines bankruptcy among its trending stories. The web crawlers used by Google News to find articles came across the story and catapulted it to prominence, resulting in a one billion dollar market loss for United Airlines, not all of which was recovered once it became clear there was no new bankruptcy. If this sort of incident can happen with Google News, then is it likely that it can happen with the fact check system, especially if it is automated just as the rest of the aggregator is? It does not seem far-fetched for an erroneous fact check containing damaging but false information about a political candidate to rise to prominence mere days before an election, all because of a mistake similar to that which caused United Airlines such severe financial harm.
Naturally, this then gives rise to the question: if Google News is responsible for swaying elections or causing businesses to lose millions because of a mistake in its automation, can it be forced to make these entities whole again? The answer is, almost unequivocally, no. First Amendment jurisprudence provides that “actual malice” would be required in order for such liability to be triggered. A burden likely too high for both the hapless politician and the now cash-strapped business to climb.
So the short answer to the title question—who fact checks the fact checkers—is that no one, and that includes Google, truly does. However, should Google News and its new fact checking tag catch on, a valuable status could be attributed to whichever fact checking sources end up most favored by the combination of algorithm and human subjectivity employed by Google in determining which fact check organizations to favor. Whether the result of this practice is the advent of a meaningful check on persistent waves of factually inaccurate journalism, more damaging mishaps—or both, remains to be seen.