Chilled Prodigy

An "anti-censorship" libel ruling threatens choice in cyberspace.

Eugene Volokh from the August/September 1995 issue

The law often demands that we sacrifice some liberty for greater security. Sometimes, though, it takes away our liberty to provide us less security. That's what a New York state court just did, and the U.S. legal liability system may have just claimed another victim: the on-line community.

In October 1994, a message appeared on a bulletin board in the Prodigy on-line service. "The end of Stratton Oakmont [a small stock brokerage] will finally come this week," the message said. "This brokerage firm headed by president and soon to be proven criminal--Daniel Porrush [sic]--will close this week." The message described a particular transaction in which Stratton was involved, and declared, "This is fraud, fraud, fraud and criminal!!!!!!!" Who wrote the message is a mystery. Though it seemed to come from a particular user's account, it turns out that someone else was using that account without permission.

Stratton naturally took umbrage, and sued Prodigy for libel, asking for $100 million in damages, plus another $100 million in punitives. And on May 24, the trial judge ruled that the lawsuit could proceed: Even though the posting, like all bulletin board postings, was made without Prodigy's prior knowledge, the judge held that Prodigy could be held liable for what it said.

Under the law of libel, it's clear that someone who republishes someone else's defamatory statement--for instance, a magazine that prints an article by an outside writer--can be held legally responsible. One can debate the wisdom of this rule, but it makes at least some sense. If my defaming you violates your legal rights, then when REASON broadcasts my defamation to 50,000 subscribers, your injury is aggravated. REASON seeks to profit from my article; REASON has the opportunity to check the article and not run it if it appears to be false; it's at least arguable that REASON should be held responsible for the injuries that its contents inflict.

But it's equally clear that people who merely redistribute defamatory statements can't be held liable. When a newsstand sells a copy of REASON, it's helping circulate its contents, including whatever nasty things I say about you; but the newsstand isn't liable because of this. Likewise, say someone gets a copy of REASON by modem. The phone company is in a sense republishing the text, but you can't get money from the company for its role in ruining your reputation. The law recognizes that no newsstand or phone company can be expected to read and verify every publication that it distributes. Imposing liability in such a situation would be both unfair and chilling to free dissemination of speech.

On-line services are decidedly closer to the phone company or the newsstand than to the magazine. Tens of thousands of messages are posted to the services daily; the services can't read them all and verify their accuracy. In the 1991 Cubby v. CompuServe case, a federal trial court recognized this and held that an on-line service couldn't be liable for defamation unless "it knew or had reason to know of the allegedly defamatory statements." Any alternative rule, the court said, "would impose an undue burden on the free flow of information."

So far, then, so good: Coming into the litigation, Prodigy could have expected that, like CompuServe, it would be immune from liability. What happened?

Well, the trouble was that Prodigy took some steps to oversee the bulletin boards that it operated. At least at one point, Prodigy used a software screening program which automatically rejected bulletin board postings that contained offensive language. Prodigy also announced to its users that they shouldn't post notes that are insulting, harassing, or in bad taste. Prodigy employees scanned the bulletin boards looking for such messages, and the Prodigy software allowed those employees to delete messages that didn't fit the guidelines. "Prodigy is committed to open debate and discussion on the bulletin boards," Prodigy said, but "this doesn't mean that 'anything goes.'"

This, the court explained, is what differentiates Prodigy from CompuServe. Prodigy controlled what was said on its bulletin boards: "That such control is not complete and is enforced both as early as the notes arrive and as late as a complaint is made does not minimize...the simple fact that Prodigy has uniquely arrogated to itself the role of determining what is proper for its members to post and read on its bulletin boards." And because of this control, for the purposes of libel law, "Prodigy is a publisher rather than a distributor." The court concluded that "Prodigy's conscious choice, to gain the benefits of editorial control, has opened it up to a greater liability than CompuServe and other computer networks that make no such choice."

The court's reasoning, though, is rather puzzling. After all, even traditional distributors, such as newsstands and bookstores, do arrogate to themselves the role of determining what's proper for their customers to read. A newsstand might choose not to carry pornography. Some bookstores refused to carry Bret Easton Ellis's novel American Psycho. Some stores specialize in libertarian books, others in communist ones. All newsstands and bookstores select their stock based on some criteria, be they popularity, topicality, nonoffensiveness (or offensiveness), or what have you.

Furthermore, the only way that Prodigy is like a magazine publisher is that it has some control over what's posted on it. But the degree of control is vastly different. I assure you that Virginia Postrel (or someone who works for her) has done more to this article than run it through a profanity-checking program. Some human being at REASON has approved every article that REASON carries--not just reserved the right to approve it or disapprove it, but has actually read it and decided whether to run it or not. Not so for Prodigy and the 60,000 messages that are posted every day to its bulletin boards.

Moreover, Prodigy doesn't purport to check messages for accuracy; it just checks them for offensiveness. It's unreasonable enough to expect Prodigy to read every message, but it simply makes no sense to expect it to fact check every message. Why should Prodigy's willingness to maintain some standards of taste translate into a responsibility for its users' factual errors?

What justification could there be, then, for this puzzling distinction between CompuServe (which does no screening) and Prodigy (which does only the limited screening that the large volume of messages allows)? Why should a small degree of extra oversight translate into a vast increase in liability?

Well, the court never directly answered these questions, but it did give one suggestion: Prodigy's screening policies themselves, in the court's view, were troublesome (even if not illegal). "It could be said," Judge Stuart Ain wrote, "that Prodigy's current system of automatic scanning, Guidelines, and Board Leaders may have a chilling effect on freedom of communication in Cyberspace, and it appears that this chilling effect is exactly what Prodigy wants, but for the legal liability that attaches to such censorship." Prodigy's willingness to restrict what's said makes it, in the court's eyes, a censor. If Prodigy does this, it must "accept the concomitant legal consequences."

It is here--in its blithe assertion that private screening and oversight is "censorship" to which there must be "legal consequences"--that the court made its greatest theoretical error. Private screening is not some shady practice employed by nasty thought controllers. Rather, it can be a positive good: It can add value to information by filtering out what people don't want to see.