To No One’s Surprise, FOSTA Is Confounding Judges–J.B. v. G6

September 13, 2021 · by Eric Goldman · in Content Regulation, Derivative Liability, Marketing

Plaintiffs underutilitized FOSTA’s new Section 230 exclusions for the first two years after the law’s enactment, but now we are getting FOSTA rulings at a rapid clip. Given FOSTA’s terrible drafting, it’s not surprising that the rulings are not agreeing with each other.

Background

As with FOSTA cases generally, this case’s facts are heartbreaking. In a prior ruling, the court held that Craigslist qualified for Section 230(c)(1) protection for state law claims but not for the 1595 claim. The plaintiff unsuccessfully asked the court to reconsider its ruling in light of Justice Thomas’ unhinged anti-Section 230 screed. Then, Craigslist asked the court to reconsider its ruling in light of Doe v. Kik and M.L. v. Craigslist. The court decides to do so and changes its mind on the 1595 claim, holding that it too is preempted by Section 230. As a result, Craigslist is dismissed from the case.

Section 230 and the 1591/1595 Interplay

FOSTA is insanely complicated, but the legal issue here is easy to state: to prove a 1595 claim, do plaintiffs also have to prove a 1591 violation? This matters because 1591 requires a higher level of defendant scienter than 1595–so high that plaintiffs will have a difficult time showing it.

Prior courts have split on this question. Doe v. Kik held plainly that 1595 plaintiffs can overcome a Section 230 defense only if they also prove a 1591 violation. However, more recently Doe v. Twitter held the opposite, saying that if Congress wanted to make it more challenging for sex trafficking plaintiffs to sue Internet services than other sex trafficking defendants, it could have said so more clearly.

This is the statutory language (from Section 230(e)(5)(A))at issue: Section 230(c)(1) does not apply to “any claim in a civil action brought under section 1595 of title 18, if the conduct underlying the claim constitutes a violation of section 1591” (emphasis added). Seeing it in cold type, perhaps you can see why I always considered this an easy statutory interpretation question. To get around Section 230, I think 1595 requires plaintiffs to prove a 1591 violation.

The court agrees based on (1) a plain reading of the text, (2) other parallel exclusions in FOSTA (where SESTA and FOSTA were grafted together), which show how Congress could have written the language differently if it wanted different results, and (3) how the totality of FOSTA gives victims a wide range of remedial relief. The plaintiff counterargued that the court’s reading wipes away the 1595 claim. The court disagrees because the plaintiff can bring it if it can establish the 1591 prerequisite, however difficult that may be. The court acknowledges that the plaintiff’s reading may make more sense as a policy result because of 1591’s higher scienter bar, but “the Court does not find that the plain language interpretation, in context, produces an absurd or unreasonable result.”

As further support for the court’s plain-reading interpretation, the court reviews the legislative history. It was painful to relive FOSTA’s tortured history, but the court does a great job recounting how things really went down.

The short story is that the initial version of SESTA didn’t link 1595 to 1591, which caused many objectors–myself included–to raise concerns that 1595’s lower scienter requirements would reach too many unintended defendants. In response to this concern, after a pivotal Senate Commerce Committee meeting (one of the most stressful moments of my professional career), the SESTA Manager’s Amendment (which got rolled verbatim into FOSTA) added the heightened scienter requirement–which is what caused the Internet Association (as ordered by Facebook) to flip on SESTA and endorse it. So efforts to write the 1591 precondition out of the statute are an attempt by plaintiffs to reclaim the original SESTA proposal, even though that’s exactly what Congress chose NOT to do as part of the bargain to advance the legislation. It’s frustrating, but hardly unexpected, to see historical revisionism by plaintiffs–but ultimately, I blame FOSTA and everyone who has supported it, because Congress was warned about these impending jurisprudential trainwrecks and went ahead anyways.

(If your head is hurting trying to sort through all of this, you are not alone).

The court show how Congress intended the deal on 1595/1591 scienter. It cites statements from Senators that they expected SESTA would have a high scienter bar:

Sen. Portman highlighted the “high bar” the “knowing standard” raises to “protect[] good tech actors and target[] rogue online traffickers like Backpage.”
“Senator Blumenthal reiterated the ‘high bar’ framing and noted ‘[i]t is time to open the courthouse doors to victims of sex trafficking who have been sold into slavery as a result of ads that right now can enjoy absolute immunity for sites that knowingly facilitate, support, or assist—knowingly facilitate, support, or assist.’”
Senator Schatz said “Congress ‘obviously want[ed] to provide space and not deter proactive actions by good actors that are doing the right thing to mitigate sex trafficking on their platforms’ and voiced the concerns of ‘big platforms’ that are ‘worried that their knowing at all triggers the knowing part of the statute.'”

The court also quoted my testimony at the Senate Commerce Committee, discussing the initial version of SESTA pre-Manager’s Amendment, where I said:

the knowledge standard, especially in the case of civil claims, is a new thing for Section 230. In general, Section 230 has not had a civil exception that has been predicated on knowledge. So the opening up of the door to looking at a site’s knowledge will be something that we haven’t seen before, and because of that we’re going to have a lot of questions. What exactly did the site know and when? And what— and how do we characterize that under the legal standards? . . .

[W]e would want to be extremely explicit about exactly when that knowledge occurred because otherwise there will be lots of discussion and debates over, well, you knew it based on you having taken this step or that step, or inferentially you should have known, or constructive knowledge, you should have been realizing what was taking place on your site. All of those will become the basis of which there will be plenty of disputes.

The court also noted my exchange with then-California AG Becerra (now head of HHS). Becerra mentioned the high scienter bar for criminal prosecutions, and I pointed out that the civil provisions in the initial SESTA version lacked that feature.

In response to the Senate Commerce Committee hearing, Senator Thune introduced the SESTA Manager’s Amendment, which modified the (e)(5)(A) provision to add the language now at issue, as well as other scienter requirements. A committee report explained that the Manager’s Amendment “introduce[d] new civil liabilities for violations of Federal criminal laws relating to sex trafficking,” implying that the 1591 criminal provisions were a part of 1595. At a second Senate Commerce Committee hearing, Reps. Wagner and Walters came to complain about the newly added scienter requirements.

Separately, Rep. Goodlatte introduced substitute FOSTA, which adopted a different structural framework in part because the 1591 scienter requirements of the SESTA Manager’s Amendment would be so hard to prove. This eventually led to the worst-of-both-worlds FOSTA where FOSTA and SESTA were combined into a Frankenstein bill. Given the survival of the SESTA Manager’s Amendment in the worst-of-both-worlds bill, the court says:

it thus appears that Congress reached a compromise by including a narrowed federal civil sex trafficking carve-out that requires plaintiffs to show the civil defendant’s knowing assistance, support or facilitation, while also enacting a criminal offense (now codified at 18 U.S.C. § 2421A) that, at least in theory, targets bad actor websites and includes a civil recovery provision.

The court then cites several floor statements from Senators referencing the 1591 prerequisites. Thus:

Plaintiff has pointed to nothing in the legislative history, even if considered, that runs counter to the Court’s plain language reading.

The court summarizes:

The Court finds that the most persuasive reading of section 230(e)(5)(A) is that it provides an exemption from immunity for a section 1595 claim if, but only if, the defendant’s conduct amounts to a violation of section 1591. To the extent Plaintiff’s interpretation is plausible, the legislative history does not present sufficiently “convincing contrary” evidence to overcome the provision’s plain meaning. Notwithstanding the policy debates aired throughout FOSTA-SESTA’s development, this interpretation is also facially consistent with the CDA’s stated policies “to preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation” and “to ensure vigorous enforcement of Federal criminal laws to deter and punish trafficking in obscenity, stalking, and harassment by means of computer.” The Court thus interprets section 230(e)(5)(A) to limit the scope of civil sex trafficking claims against interactive computer services that otherwise meet the requirements for CDA immunity to circumstances in which the defendant’s conduct amounts to a violation of section 1591

The plaintiff never attempted to satisfy the 1591 requirements, so the court grants Craigslist’s motion to dismiss the 1595 claim.

Given the collision between this ruling and the Doe v. Twitter case, and the odds that both of these cases will go to the Ninth Circuit, we will have to wait for an appellate ruling before we can more confidently predict what the next court will do.

Case Citation: J.B. v. G6 Hospitality, LLC, 2021 WL 4079207 (N.D. Cal. Sept. 8, 2021)

More SESTA/FOSTA-Related Posts:

To No One’s Surprise, FOSTA Is Confounding Judges–J.B. v. G6

Comments and Pings