At the dawn of the Internet’s emergence, the Supreme Court rhapsodized about its potential as a tool for free expression and political liberation. In ACLU v. Reno (1997), the Supreme Court adopted a bold vision of Internet expression to strike down a federal law - the Communications Decency Act - that restricted digital expression to forms that were merely “decent.” Far more than the printing press, the Court explained, the mid-90s Internet enabled anyone to become a town crier. Communication no longer required the permission of powerful entities. With a network connection, the powerless had as much luck reaching a mass audience as the powerful. The “special justifications or regulation of the broadcast media” had no application to the “vast democratic forums of the Internet.”

Twenty years later, the Roberts Court had an opportunity to explain how the First Amendment should operate in the mature Internet of 2017. Despite the interval of time, the Roberts Court of 2017 took a remarkably similar approach to the Rehnquist Court of 1997. In Packingham v. North Carolina, Justice Kennedy announced the start of the “Cyber Age.” The Internet was the virtual public square, much like streets and parks. Because the “Internet” was still in its infancy, its impact on expression was not fully understood. The expressive potential of the “Internet” would be imperiled in the absence of a hands-off approach. Justice Kennedy noted that someday, the Internet might be used for anti-social ends. Until then, extreme caution was in order so the Internet’s democratic potential could be realized.

Contrary to the Court’s thinking, the Internet is no longer in its infancy. It has matured at a breathtaking pace. Virtually all aspects of our public and private lives - politics, child-rearing, work, health, shopping, and sex - involve the Internet. If online discourse ever accorded with the Court’s vision, it does not now. Rather than just the virtual town square, the “Internet” is bound up in everything and everywhere-whether the workplace, library, coffee shop, gym, park, public street, town square, or bedroom.

This article debunks the Court’s magical thinking about the Internet. The Internet’s expressive opportunities are not available to all on equal terms, thanks to the wide availability of personal data. Online platforms highlight favored content while burying disfavored ones. Search engines produce different, and less advantageous, results to people of color and women than to men. Cyber mobs shove people offline with doxxing, swatting, and other privacy-invasive forms of abuse. Online platforms fuel polarization and filter bubbles, ensuring an electorate without access to a full range of ideas and information.

Fake news spreads like wildfire on social media platforms that are often people’s main source of information. We need clear principles to guide and secure meaningful digital free expression. This article charts a path to provide just that. Part I exposes crucial myths surrounding the digital speech and privacy in our networked age. Part II offers a conception of free speech based on a distrust of power, both public and private. Even if doctrinal analysis does not account for private barriers to free expression, the project of free expression should. Part III lays out four essential preconditions for a theory and a system of free expression in the digital age. These preconditions are substantive and procedural. They require legal intervention and extra-legal efforts. They draw some inspiration from due process guarantees and some from commitments to equality. Underlying these principles is a unifying normative commitment: If we want to ensure that our commitment to long-standing democratic theories of free expression survives its translation to the digital environment, we need to take a long, hard look at the digital public sphere we actually have, rather than one that we might want or one that has been advertised to us by Silicon Valley.

Danielle Citron & Neil M. Richards, Four Principles for Digital Expression (You Won’t Believe #3!), 95 Washington University Law Review, 1353–1388 (2018).