Jeff Kosseff

For me, at least, this sophomore effort does not rise to quite the high level of his most excellent The Twenty-Six Words The Created the Internet. Still, it is full of good information and perspective.

I don't agree with many of his policy recommendations (truth be told, he's too much of a libertarian for my taste, and thus he tends to be willing to tolerate a great deal of life warping trolling on the Internet to allow for the possibility that there might be a Thomas Paine out there somewhere, who needs someday to publish a blog anonymously). Regardless, he makes sincere efforts to defend his positions, which make them interesting foils to alternative viewpoints.… (mehr)

The author offers plenty of original research into the background of CDA Section 230, and of many of the important cases that helped to determine its meaning. All in all, a good story, well told. The final chapters were a bit self-centered I thought, but perhaps he's entitled to that.

Having argued that 230 built the Internet we see today, he surprisingly advocates for changing it to include more exemptions to its application. We have a president demanding that ISPs that don't flatter him should be sued, so the author may get his wish.… (mehr)

A history of Section 230 of the Communications Decency Act, which protects website owners from liability from stuff that their users post. (Intellectual property claims, child porn, and violations of federal criminal laws are not covered by 230.) Kosseff traces the legal background in the US that produced the perceived need for 230—courts had started to suggest that websites could be liable for user-supplied content if they enforced content policies but missed a particular bad post, but not if they didn’t check content at all, giving them an incentive to be completely hands-off. Congress wanted to encourage sites to monitor instead of having a free-for-all. 230 passed as part of an overall attempt to get porn offline, and Kosseff points out that 230 only made it in as a compromise with the porn-regulators; after the porn regulations in the CDA were struck down, only 230 remained, and it made American companies unstoppable in terms of what they were willing to experiment with, for good and ill.

Kosseff tracks the early breadth of 230 cases, protecting site operators against liability for user-supplied defamation and also blocking causes of action that tried to dodge 230 by pleading that the legal violation at issue wasn’t based on users’ content. (It usually was.) Over time, however, litigants have figured out some paths that occasionally work, based on theories such as a breach of a promise from the site operator or the operator’s involvement in developing or enhancing specifically illegal content. It’s a good overview.

I have some serious disagreements with how Kosseff approaches the recent modification of 230 to create more liability for involvement in sex trafficking—begun under Obama but signed into law by Trump. The problem was Backpage, and Kosseff highlights horrible stories of girls who were coerced and raped and whose profiles appeared on Backpage. “Viewing the reports in the most charitable light, it is clear that Backpage knew that its users were posting sex trafficking ads, yet it failed to take all possible steps to stop them.” But litigation against Backpage seemed stalled, in part because the litigants used bad theories/didn’t have the strongest evidence that came out later. Kosseff criticizes the major internet companies for testifying that these stories were bad, but that modifying 230 would “create a new wave of frivolous and unpredictable actions against legitimate companies rather than addressing underlying criminal behavior.” He says: “Such a statement might be appropriate for, say, data security legislation. But this is about children being raped and murdered.” I find this response close to immoral itself: when Trump talks about immigrants being rapists, or when homophobes do the same with bathroom bills, we do not say “you can’t trivialize rape by saying our solution is wrong!” Targeting internet services is not discriminatory, but it’s also not a good idea, and I think we should get to point that out.

Indeed, Kosseff himself says “[e]ven if public sites such as Backpage were shut down, I have no doubt that the pimps would continue to shift to the darker corners of the web.” But he’s still “appalled by the technology companies’ general approach to the issue” because they didn’t “provide viable alternative solutions that would allow states to prosecute and victims to sue the websites that knew about, encouraged, or even participated in the sale of children for sex.” This is so even though he believed that Backpage’s specific actions deprived it of 230 immunity because it participated in the development of trafficking posts specifically; he just thought “we can’t wait for the courts to get it right.”

But… he doesn’t like the law that was ultimately adopted, because it’s overbroad and threatens services with liability even if they didn’t specifically know about particular instances of trafficking. Which is exactly what the services were saying. Worse, as he points out, the law had no impact on Backpage itself, because it was promptly shut down for violating federal criminal law based on pre-amendment law. The change was thus unnecessary to get Backpage, but it did succeed in shutting down Craigslist personals and making it incredibly risky to operate any service for voluntary sex workers. There’s not a word in the book about sex workers’ concerns about losing the ability to screen clients and avoid unsafe situations on the street if online venues shuttered—even though those predictions do seem to have come true.

So by Kosseff’s own account, (1) changing 230 wasn’t necessary, (2) changing 230 won’t work against dedicated traffickers, and (3) the change “is ambiguous and overbroad and leaves well-intentioned platforms with the choice of censoring legitimate speech or risking lawsuits and criminal prosecution,” but it’s still (4) it’s services’ own fault because they didn’t take victims seriously enough. As we’ve seen with other disputes over content regulation online, if you’re not perfect, you can’t escape criticism, and at internet scale it’s impossible to be perfect. American free speech exceptionalism is not always a good thing, but it often looks a lot better than the alternatives, especially with a litigation culture that leads us to sue more—and ask for and get bigger damage awards than are available in other countries, which is an underappreciated part of the stakes in the US—than people do in other countries.… (mehr)