The Eliminating Abusive and Rampant Neglect of Interactive Technologies Act of 2020, or EARN IT Act, is set for a markup in the Senate Judiciary Committee as early as this Thursday, June 25. Essentially the bill conditions intermediary liability protections for web services, known as Section 230, on yet-to-be-determined regulatory guidance regarding online child sexual abuse material (CSAM). While the need to combat CSAM is clear, the EARN IT Act is a fundamentally flawed vehicle for doing so.
As the name implies, the EARN IT Act’s sponsors suggest online services must prove themselves deserving of Section 230 protections. For those unfamiliar, Section 230 is a law that says, for the most part, that you are responsible for what you post online, not the web service itself. This assignment of legal liability has allowed web services to filter out harmful content, without necessarily assuming liability for anything they may miss. Hence the rise of the Internet as we know it today, from Airbnb to Zoom.
Most recent proposals to reform Section 230 hinge on the idea that Section 230 is some sort of unique subsidy for web services, and therefore it is reasonable for the government to attach strings to it to accomplish other policy ends. The EARN IT Act is obviously no different.
Yet Section 230 is not the unique privilege the detractors of tech companies would have you believe, and the idea that companies should have to “earn” it is simply egregious.
All Section 230 does is place liability where it logically belongs. Web services like Google, Facebook, Ebay, and other sites where third-party content is key to the business are subject to the network effect—they generate more value as more people use them. Users derive more value as they are offered more content and connections and third-party sellers and advertisers benefit from exposure to more eyes and ears. So, these web services necessarily function best for most at enormous scale. That enormous scale would be virtually impossible without Section 230.
Unlike book publishers or newspapers, to which they are often compared by Section 230 critics, web services simply cannot prescreen the thousands, millions, or even billions of posts users uploaded to their sites on a daily basis and still function at the speed users have come to expect.
Furthermore, it is impossible for content moderators to know the context of many posts. What looks like slander may just be a joke. Post-hoc content moderation even at a small scale is difficult at best but is integral to how these businesses function. Making companies liable for any and every needle they may miss in the constantly growing haystack of online content is completely unreasonable.
Yet that was where the law stood prior to Section 230. Web services faced what is known as the moderator’s dilemma. If they moderated user-generated content on their sites, they assumed liability for all of it. This created an incentive not to moderate. A world where Facebook and YouTube are filled with pornography, violence, spam, and other objectionable content that is otherwise perfectly legal is not a world in which anyone wants to live.
Imagine such a liability standard placed on any other industry. For example, imagine if automakers assumed liability for the actions of every driver and the results of any and all traffic accidents involving their vehicles if they decided to improve the safety features of their vehicles. Would that make sense to anyone? It doesn’t make sense to hold automobile companies liable for the actions of a driver in any situation.
If product manufacturers could be held liable for any and all actions of the users of their products, regardless of their knowledge, then store shelves would be rather empty. So why should web services have to earn what is essentially common sense for other industries?
Even if one still considers web services to be analogous to traditional publishers—which again they are not, since traditional publishers must affirmatively decide whether to publish something before it is seen—the trajectory of publisher liability case law suggests Section 230 is no unique deviation. Brent Skorup and Jennifer Huddleston recently wrote the following in an article published in the Oklahoma Law Review:
The Section 230 reform movement is growing, and many of the reform arguments complain that online intermediaries receive a special dispensation regarding publisher liability. The truth is more complicated. Starting in 1931 and for six subsequent decades, courts gradually chipped away the regime of strict liability for publishers and content distributors owing to the practical difficulties of screening all tortious content and to the potential for restricting First Amendment rights. Those courts found that mass media distributors warranted extensive liability protections, including an important protection for conduit liability.
Skorup and Huddleston’s argument raises another key point about why suggesting Section 230 is something to be earned is so troubling. Beyond a common-sense application of liability, where it properly lies, Section 230 protects companies as they exert their First Amendment-protected rights of speech and association.
As mentioned, the moderator’s dilemma that existed prior to Section 230 left web services in a position where deleting or otherwise moderating content on their websites could leave them liable for all content on their sites, even content of which they were unaware. However, companies don’t have to host or otherwise affiliate with content they find objectionable. They are free to moderate such content as they see fit under the First Amendment. This principle was recently affirmed in a unanimous decision handed down by the Ninth Circuit Court of Appeals.
In short, in a practical sense, Section 230 is the guarantor of web services’ ability to exercise their constitutionally protected rights. Yet, the EARN IT Act implicitly suggests that web services must earn the ability to exercise their rights. That’s not how rights work.
There are a number of other problems with the EARN IT Act. CEI has warned about these before. But the message such legislation sends—the notion that a private entity such as a web service must first prove itself worthy of logical treatment and having its rights protected by government—is perhaps its most alarming aspect.