THE DUTY NOT TO CONTINUE DISTRIBUTING YOUR OWN LIBELS.

• Author: Volokh, Eugene

INTRODUCTION 316 I. REASONS FOR THE DUTY 319 II. THE CONSTITUTIONALITY OF THE DUTY 323 III. THE EXISTING DUTY NOT TO CONTINUE DISPLAYING POSTS ON PHYSICAL PROPERTY 325 IV. SCOPE OF THE DUTY: NEGLIGENCE-BASED LIBEL CLAIMS 327 V. THE SINGLE PUBLICATION RULE 329 VI. PRACTICAL APPLICATIONS AND PRACTICAL PROBLEMS. 332 A. Immediate Demand from Plaintiff 332 B. Years-Later Demand from Plaintiff 333 C. Months-Later Demand from Plaintiff 334 D. Repeated Demands from Plaintiff 335 E. How Long Should Defendant Have for the Renewed Investigation? 336 VII. SUMMARY: WHAT NEEDS TO BE CORRECTED? 337 A. No Duty to Remove or Correct Third Parties' Posts. 337 B. Duty Arises Only When the Plaintiff Has Given the Defendant Notice 338 C. Duty Arises Only When the Defendant Can Remove or Modify the Material. 338 1. Third-party quotations of defendant's material 338 2. Material that the authors can't edit. 339 3. Images, videos, and audio recordings 340 4. Defendants with lost privileges 340 D. Duty to Promptly Remove or Modify 341 E. Not a Duty to Publish a Retraction 341 VIII. A SAMPLE STATUTE 341 IX. DUTY TO STOP REPORTING HIGHLY INCOMPLETE REPORTS OF LEGAL PROCEEDINGS 343 A. Knowingly Incomplete Reports as Libel 343 B. Extending the Statute of Limitations for Legal Updates 346 CONCLUSION 349 INTRODUCTION

Donna writes something false online about Paul--sincerely, even reasonably, believing it to be true. This absence of a mens rea keeps her from being liable for defamation.

Paul then promptly tells Donna that her post is false, and backs that with persuasive evidence; maybe it's as simple as a case of mistaken identity. Yet Donna continues to keep her article (or blog post or social media post) online, now with "actual malice"--knowledge that the statement was false, or at least recklessness about that possibility. Every day, the article is distributed to more readers, for instances ones that find it via a Google search.

Should Donna (and her employer (1)) be liable for defaming Paul, based on the continued distribution, even though she is not liable for the initial publication? Or should she be immune from liability, even if she keeps the material up unmodified, because she wasn't culpable at the time she made the statement? The answers, surprisingly, are unsettled.

Or say Donna accurately posts online that Paul has been convicted of a crime. Her statement, and her description of Paul's actions, fairly and accurately reports on government proceedings, and is thus not libelous. Three months later, the conviction is reversed because there was insufficient evidence supporting it.

Paul informs Donna about that. Donna is now knowingly distributing an article online that no longer fairly and accurately summarizes the aggregate of the legal proceedings in the case. (As we'll see in Part IX, reports that mention a conviction without mentioning the reversal are generally not covered by the fair report privilege.) Should Donna be liable for continuing to knowingly distribute the now-defamatory material? (2) The answer is likewise unsettled.

The internet is a persistent medium, where defamation often causes damages through a steady drip-drip-drip of people finding items online each day, rather than through the short, sharp shock of a traditional print publication. But our libel law developed when publishers printed something and it then left their control. In such situations, the only questions were whether the publishers were liable for the initial printing, and perhaps whether they should have an affirmative duty to publish a retraction. The question whether they should have a duty as to material that they were continuing to distribute rarely arose. (3) Yet that question is especially important today.

In this Article, I discuss such liability for continuing to distribute material once one knows it's libelous, and tentatively argue that there should indeed be such liability (properly bounded):

* It's fair both to publishers and to the victims of the false statements. (4)

* It's consistent with First Amendment principles. (5)

* It's authorized by a longstanding libel law principle applicable to real property owners, who can be liable for continuing to keep defamatory material on their property once they learn of its presence."

* It should apply in some measure to private-figure libel compensatory damages cases that are based on the defendant's negligence, and not just to cases that are based on the defendant's "actual malice." (7)

* And such liability shouldn't be seen as contrary to the "single publication rule," properly understood. (8)

Such liability does impose some burden on those who have posted the statements, and that gives me pause. But that burden strikes me as on balance justifiable. It's not far from the normal burden that modern libel law--sharply constrained by the First Amendment--generally imposes on speakers. (9) And it's suitably limited to scenarios where authors and publishers are practically able to remove or correct material that they have been informed is likely mistaken. (10) Courts should recognize it under existing common-law principles, and state legislatures may institute it, too; I offer a sample statute in Part VIII.

The liability should also apply, as I suggested above, to situations where a legally significant decision that strongly implies likely guilt is reversed, for instance when a prosecution leads to an acquittal, or a conviction is reversed on the merits. In that situation, publishers should have a duty not to continue hosting material that has become misleadingly incomplete in important ways, though they should be free to keep the original report up with an update indicating what later happened. (11) And for these particular legal updates (unlike the other updates I discuss above), the statute of limitations should be extended. (12)

There are three important limitations to my claims here:

1. I'm not speaking of a "duty to retract," in the sense of a print newspaper's affirmative obligation to publish a retraction to a printed story when it learns of errors. Such a duty has generally been rejected as a matter of libel law, (13) and I am not trying to revive it here. American common law generally (though not uniformly) disfavors such affirmative duties, and I don't seek to refight that battle.

   Rather, I'm speaking here of a negative duty, closely linked to the traditionally recognized duty not to defame people in the first place--the duty to stop defaming someone, by removing or correcting online material over which one has control.

2. Potential liability would be triggered only when the subject of an article notifies the publisher that the article is in error (or that there have been new legal developments that essentially overturn an earlier legal determination of probable guilt). Publishers wouldn't have to proactively do follow-up investigations in the absence of such notifications. (14)

3. I am speaking here of a publisher's duty to stop distributing its own errors. I am not proposing changing 47 U.S.C. [section] 230(c)(1), which limits publishers' responsibility for material submitted by others (such as comments posted by readers below news stories or blog posts). There is a separate debate about whether [section] 230(c)(1) should be modified; I leave that to other articles.

   1. REASONS FOR THE DUTY

   Let's turn to a more concrete hypothetical: say two reporters, Ophelia Often (who tends to check her voicemail often) and Randy Rarely (who tends to check his rarely), are writing stories about Starlight Rainbow, (15) accusing her of mistreating a fifth-grade student. (For convenience, assume that Starlight is a principal and thus a public figure or public official under state law. (16)) It turns out, though, that both reporters erred: the actual allegations of mistreatment were about a different teacher with the same last name, Cynthia Rainbow. (17)

   Starlight learns about the planned stories, and leaves voicemails for both reporters with persuasive evidence that she's not actually the guilty party. (She actually works at a different school.) Ophelia listens to her voicemail before her story is posted, but Randy listens to his only after. For whatever reason, Ophelia still posts her story, and Randy doesn't correct the story he had posted. (18)

   Starlight now sues Ophelia (and her employer) for posting her story and Randy (and his employer) for continuing to keep his story up. The statement in each story--that Starlight was accused of mistreating the student--is false and defamatory. Ophelia and Randy are both aware now that it's probably false. Both employers are keeping up the stories without correction, even though they are aware that the stories contain false and defamatory statements.

   Starlight's claim against Ophelia's employer will thus likely prevail: Ophelia posted knowing that the statement was probably false (which likely counts as "reckless disregard" of the truth and therefore "actual malice" (19)), and liability is imputed to Ophelia's employer under respondeat superior. (20) Starlight can thus use the threat of liability to pressure Ophelia's employer to correct the story on its site. And it's hard to see why Starlight's claim against Randy should be treated any differently:

1. The harm caused by the stories is identical: Starlight is being damaged equally by both. (21)

2. The value of the statements about Starlight is equally low in both stories: both statements are false. (22)

3. The current mental state of the reporters and employers is equal: Randy and Randy's employer are as aware of the falsehood now as Ophelia and Ophelia's employer were when Ophelia's story went up.

4. The current culpability of the reporters and employers is thus also equal: Randy and Ophelia are continuing to distribute material that they now know to be false, which is culpable whether or not their initial posting was culpable at the outset (as Ophelia's was but Randy's wasn't).

5. The chilling effect...