A question arose in connection with the Brandon Darby / New York Times controversy: Say that a newspaper (or some other entity) posts its own article on its site, believing it to be true, and not having serious doubts about its truth. At that point, even if some allegations in the article are actually false, the publisher has not libeled any public figure mentioned in the article, because the publisher was not acting with what the law calls “actual malice” — knowledge of falsity, or reckless disregard of a known serious risk of falsity. Likewise, the publisher has not libeled any private figure mentioned in the article, if the publisher reasonably believes the article to be true, because the publisher was not acting with negligence (the standard for private figures). (Note: This is an oversimplification, but sufficient for our purposes.)
But say that the publisher later learns that the statements are false. Is it liable for failing to remove the allegations, at that point, on the theory that it is now acting with “actual malice” (or negligently, if the plaintiff is a private figure)?
The answer appears to be yes. The Restatement (Second) of Torts — not a statute, but an influential summary of court decisions — announces this general principle in § 577(2):
One who intentionally and unreasonably fails to remove defamatory matter that he knows to be exhibited on land or chattels in his possession or under his control is subject to liability for its continued publication.
Comment: ... p.... The basis of the liability is his duty not to permit the use of his land or chattels for a purpose damaging to others outside of the land.... So far as the cases thus far decided indicate, the duty arises only when the defendant knows that the defamatory matter is being exhibited on his land or chattels, and he is under no duty to police them or to make inquiry as to whether such a use is being made. He is required only to exercise reasonable care to abate the defamation, and he need not take steps that are unreasonable if the burden of the measures outweighs the harm to the plaintiff. In extreme cases, as when, for example, the defamatory matter might be carved in stone in letters a foot deep, it is possible that the defendant may not be required to take any action at all. But when, by measures not unduly difficult or onerous, he may easily remove the defamation, he may be found liable if he intentionally fails to remove it.
Illustration: 15. A writes on the wall of the men’s washroom in B’s tavern a statement that C is an unchaste woman. B fails to discover the writing for an hour. After he discovers it, he fails to remove it for another hour, although he has ample opportunity to do so. During the second hour the writing is read by several men. B is subject to liability for the continued publication of the libel during the second hour, although not for the original publication. [There's an actual California court case on that. -EV]
The logic of this provision seems to apply to Web sites owned by the publisher, especially when they are on servers owned by the publisher (since such servers would be the publisher’s chattels), though I think it would also extend even to the Web sites as a form of intangible chattel. (The Restatement is a summary of court cases, not a statute, so one can apply it by analogy even to situations that may fall outside its literal words.) A 2007 federal district court case has indeed applied this principle to Web sites, and a 2008 South Carolina case seems to have endorsed this principle as to Web sites, but concluded it was inapplicable on the facts (since the failure to remove was only negligent and not deliberate). I know of no case that has rejected it as to Web sites. For an opinion on a related issue by a Texas appellate court — the Darby case has been filed in Texas and will likely be decided under Texas law (since Darby is a Texas resident) — see this case. The opinion is a dissent, but the majority simply didn’t discuss this theory, rather than rejecting it expressly.
Note that this principle only applies to the publisher’s own stories. If the story is not the publisher’s own, then the publisher will likely be be immune under 47 U.S.C. § 230. And again, note that I don’t know what the facts are in the Darby / Times controversy; I am just setting forth the likely general legal principle.
Note also that the publisher will generally not be liable once the statute of limitations (generally a year or longer) has run since the original post. At that point, under the “single publication rule” — which is generally accepted in most states, and has generally been applied to the Internet in the cases that have considered the issue — no further lawsuits can be brought based on the original post, even if the publisher eventually learns that the post is false. [UPDATE: A commenter was confused by my initial, briefer treatment of this issue, so I added this paragraph and moved my briefer reference to the statute of limitations out of an earlier paragraph.]
You might ask: Does this mean that traditional libraries and booksellers had a duty to remove libelous material from their stacks and shelves, once they were aware that it is libelous (assuming there’s no statute of limitations problem, of the sort discussed in the preceding paragraph)? The answer appears to be yes as to booksellers, see Spence v. Flynt, 647 F. Supp. 1266, 1274 (D. Wyo. 1986) (which applies well-established though rarely litigated rules related to “distributor liability” for defamation), and logically this suggests the same result would be true as to libraries. There are plausible policy arguments in either direction on this, so perhaps courts might have reached a different result as to libraries; but at least a credible claim could have been made against them. Finally, note that Lexis and similar services would not have such a duty, but only because of the federal immunity for Internet distributors secured by 47 U.S.C. § 230.