Censorship by proxy: the First Amendment, Internet intermediaries, and the problem of the weakest link.

• Author: Kreimer, Seth F.

The rise of the Internet has changed the First Amendment drama, for governments confront technical and political obstacles to sanctioning either speakers or listeners in cyberspace. Faced with these challenges, regulators have fallen back on alternatives, predicated on the fact that, in contrast to the usual free expression scenario, the Internet is not dyadic. The Internet's resistance to direct regulation of speakers and listeners rests on a complex chain of connections, and emerging regulatory mechanisms have begun to focus on the weak links in that chain. Rather than attacking speakers or listeners directly, governments have sought to enlist private actors within the chain as proxy censors to control the flow of information.

Some commentators have celebrated such indirect methods of governmental control as salutary responses to threatening cyberanarchy. This Article takes a more jaundiced view of these developments: I begin by mapping the ubiquity of efforts to enlist Internet intermediaries as proxy censors. I emphasize the dangers to free expression that are likely to arise from attempts to target weak links in the chain of Internet communications and cast doubt on the claim that market mechanisms can be relied upon to dispel them. I then proceed to explore the doctrinal resources that can meet those dangers.

The gambit of enlisting the private sector to establish a system to control expression is not new in the United States. I argue that the First Amendment doctrines developed in response to the last such focused effort, during the McCarthy era, provide a series of useful starting points for a First Amendment doctrine to protect the weak links of the Internet.

INTRODUCTION I. THE PHENOMENON: PROXY CENSORSHIP AND INTERNET INTERMEDIARIES A. Proxy Censorship Abroad B. Proxy Censorship in the United States II. FREE EXPRESSION AND THE PROBLEM OF THE WEAKEST LINK A. The Dangers of Proxy Censorship. B. The Coasian Counter and Its Limits. III. FIRST AMENDMENT DOCTRINE AND THE PROBLEM OF PROXY CENSORSHIP A. Learning from History: The McCarthy Era, Indirect Sanctions, and the Suppression of Dissent B. Doctrinal Responses to Indirect Sanctions 1. "Subtle Government Interference": Indirect Censorship as Constitutional Violation 2. The Problem of Chilled Intermediaries in Old Media and New a. Print, Film, and Broadcast Intermediaries b. New Media and the Problem of Chilled Intermediaries i. Video Recorders: The Manufacturer as Intermediary ii. The Cable Trilogy: The Danger of Networks as Proxy Censors iii. Subtle Interference and Internet Intermediaries C. Doctrinal Structures To Address the Problem of the Weakest Link 1. The Doctrinal Heritage 2. Collateral Damage Doctrines and the Problem of Proxy Censorship of the Internet a. Precision of Regulation and Collateral Damage b. "Less Intrusive Alternatives". D. Safe Harbors and Clear Boundaries: The Danger of Liability Without Fault or Falsity 1. The Doctrinal Heritage a. First Amendment Skepticism of Strict and Vicarious Liability b. Transmission of Truth and Constitutional Privilege 2. Fault, Falsity, and the Problem of Proxy Censorship of the Internet a. Vicarious Liability for Copyright Violation b. Material Support, the "War on Terror, "and the Internet c. Safe Harbors Beyond Intent: Privilege for Weaving the Internet CONCLUSION INTRODUCTION

The archetypal actors in the First Amendment drama appear on stage in dyads: in free speech narratives, a speaker exhorts a listener; in free press accounts, a publisher distributes literature to readers. (1) In the usual plot, the government seeks to disrupt this dyad (for legitimate or illegitimate reasons) by focusing sanctions on the source of the speech. The government attempts to license her, to tax her enterprise, or to threaten her with civil or criminal penalties; courts respond by evaluating the legality of those attempts. On occasion, the government turns its efforts to the listener, seeking to punish receipt of illicit messages or possession of illicit materials preparatory to reading them, and the courts proceed to evaluate the constitutionality of those proposed sanctions.

The Internet, as a network of networks, alters the drama. When communication utilizes the Internet, government finds it more difficult to sanction either speaker or listener. Speakers can hide their identities, impeding direct coercion; they can extend the reach of their communications into foreign jurisdictions that may face legal or practical impediments to exerting control. Even where speakers are theoretically subject to sanctions, the exponential increase in the number of speakers with potential access to broad audiences multiplies the challenge for censors seeking to suppress a message. (2) On the listeners' side, an expanding universe of seekers of forbidden content can obtain access to material in private without leaving their homes, bypassing both formal and informal obstacles, and can pursue alternative pathways when a particular route is blocked. The mantra is that "the Internet interprets censorship as damage, and routes around it." (3)

Faced with these challenges, state actors who seek to control Internet communications have begun to explore strategies that target neither speakers nor listeners. Regulators have fallen back on alternatives predicated on the fact that, in contrast to the usual free expression drama, the Internet is not dyadic. The Internet's resistance to direct regulation of speakers and listeners rests on a complex chain of connections, and emerging regulatory mechanisms have begun to focus on the weak links in that chain. Rather than attacking speakers or listeners directly, governments have sought to enlist private actors within the chain as proxy censors to control the flow of information.

Some commentators have celebrated such indirect methods as salutary responses to threatening cyberanarchy. Jack Goldsmith opines that local control of service providers could allow governments appropriate leverage over foreign content, (4) while Neal Katyal argues that Internet service providers (ISPs) "will often be essential in preventing cybercrime." (5) Jonathan Zittrain has taken the position that judicious pressure on Internet "points of control" offers an admirable "chance of approximating the legal and practical frameworks by which sovereigns currently sanction illegal content apart from the Internet," (6) while Joel Reidenberg commends the enlistment of Internet intermediaries as a means of reestablishing the primacy of "democratically chosen law" and the ability of states to protect their citizens. (7) Ronald Mann and Seth Belzley laud the promise of exerting control over payment intermediaries, (8) while Doug Lichtman and Eric Posner have argued strenuously that "ISPs should be called into the service of the law" by imposing vicarious liability. (9)

This Article takes a more jaundiced view of these developments. I begin by mapping the ubiquity of efforts to enlist Internet intermediaries as proxy censors. I emphasize the dangers to free expression that are likely to arise from attempts to target weak links in the chain of Internet communications and cast doubt on the claim that market mechanisms can be relied upon to dispel them. I then proceed to explore the doctrinal resources by which a system of free expression can meet those dangers.

The gambit of enlisting the private sector to establish a system to control expression is not new in the United States. I argue that the First Amendment doctrines developed in response to the last such focused effort, during the McCarthy era, provide a series of useful starting points for a First Amendment doctrine to protect the weak links of the Internet.

1. THE PHENOMENON: PROXY CENSORSHIP AND INTERNET INTERMEDIARIES

   The very plurality of private actors who cooperate to achieve Internet communications provides governments seeking to recruit proxy censors with a target-rich environment in three dimensions. (10) First, the networks of the Internet involve a series of electronic links; at each link, from user to originating computer to server to ISP to Internet backbone and back down the chain to the end user, the state may find a potential proxy censor. Each intermediary may interdict communications, or identify speakers, listeners, or other intermediaries against whom sanctions may be directed. (11)

   Second, as Herbert Simon pointed out a generation ago, "a wealth of information creates a poverty of attention," (12) and the wealth of information on the Internet multiplies at an exponential rate. The emerging organization of Internet communications relies on search engines, blogrolls, RSS feeds, links, and other directories to allow users to sort through the vast amounts of available information and allocate their limited stock of attention. These actors constitute a second set of potential weak links between speaker and listener. Even where a prospective speaker retains the technical capacity to reach prospective listeners, she must still catch their attention in a communicative universe populated with upwards of 600 billion webpages and 50 million blogs. (13) A government intervention that interferes effectively with the ability of intermediaries (whether search engines or well-attended websites) (14) to direct attention to particular speakers renders those speakers as unable to communicate with mass audiences as if they were silent and invisible.

   The technically demanding structure of the Internet offers a third set of potential weak links, in the form of the providers of specialized equipment and services upon which effective Internet access depends. To the extent that potential regulators can induce these providers to disrupt communications, whether by blocking payment to targeted websites, or by embedding obstacles to communication and mechanisms of surveillance in the hardware or software that facilitates communication, they can spawn effective proxy...