Yesterday afternoon, which is prime “news dump” time, I published some thoughts on social media’s prospects for the future–whether they hinge on the adoption of a “decentralized” protocol, or if some changes in the law might create a more hospitable environment for those who want to offer a platform which respects free thought and expression, and user privacy, while providing avenues for meaningful monetization. Check it out over on my Substack here, and please subscribe there if you haven’t yet!
Over the last week I’ve been working with my colleagues to rebuild Parler’s infrastructure, improve our guidelines enforcement process, and get back online. (Step one: a static web page.) I’ve also been making our case in the media—and in doing so have been mentally processing the injustice that has been done both to Parler, and to those who have relied upon Parler to be able to express themselves online. Thanks to those who have invited us to make our case, as well as to those—like the authors of this excellent opinion piece—who have contributed to my thinking on the relevant issues. (Standard Disclaimer: I speak only for myself here, and any errors in presentation or inference are mine.)
There are many staunch defenders of Section 230, which grants legal immunity to platforms for user-generated content, as well as for “good faith” decisions to remove or otherwise curate “objectionable” content. The above-linked piece (link again here) calls into question the wisdom of this immunity, at least when it exists alongside pressure on private companies from legislators to remove content from their platforms, when that content would otherwise be protected by the First Amendment. The authors cite legal precedent holding that conduct of “private” companies for which government grants immunity, and which government pressures them to engage in, is better thought of as publicly enforced conduct of a private company. And so, while many of us (I was included) have resisted referring to content moderation by private companies as “censorship,” we might need to consider calling it “censorship-by-proxy.”
Now recall that Mark Zuckerberg, in the most recent Big-Tech-CEO-Hearanguing before Congress, suggested amending Section 230 as follows:
“Transparency” – each company enjoying Section 230 immunity would be required to issue periodic reports detailing how it dealt with certain types of “objectionable” content.
“Accountability” –platforms enjoying immunity could also be held to some minimum level of “effectiveness” with respect to dealing with that “objectionable” content. (Recall he also bragged about how effective Facebook’s “hate speech” algorithms are.)
Perhaps you think “transparency” at least, is good. But imagine what information ends up being collected and retained as “ordinary business records” when complying with this sort of law, and read on.
In the last week-plus, we’ve seen a chorus of people blaming Parler, specifically, for threats or incitement in user-generated content. According to these voices, Parler’s not dealing with this content adequately was responsible for the inexcusable actions of a number of individuals on January 6. Setting aside issues of free will, consider the fuller factual picture that has since been revealed: Parler’s competitors’ platforms were also filled with this content, and some blame Facebook for playing a much larger role in facilitating the planning that led up to the 6th.
Yes, that’s a Salon article. What does Salon hope to gain by blaming Facebook and showing sympathy to Parler? I argue that placing responsibility for user-generated content on platforms plays right into the totalitarians’ hands.
With all the platforms now being blamed for user-generated content containing threats or incitement, the new Congress needs only to accept Mark Zuckerberg’s engraved invitation to amend Section 230 along the above lines. But, as we’ve learned in the last week, no system of guidelines enforcement is perfect. If Facebook, with all its algorithms and other resources could not “adequately” deal with this content, then what company could?
If it’s not actually possible to be good at this, to the standard that everyone seems to expect, and Zuckerberg is calling for all of us to be regulated according to that standard, then what exactly is he calling for (whether he realizes it or not)? For government to take over, to have arbitrary control. For all online platforms to operate only by permission of government, according to whatever standards politicians (or the Twitter mobs pulling their strings) deem fit—and this will be true with respect to both free speech and privacy.
As for free speech, not only has Zuckerberg invited “hate speech” regulation, I’ve also learned this week that the leading third-party AI solutions seem to be much better at detecting “hate speech” than they are at detecting threats or incitement. Perhaps this is because many platforms have elected to moderate “hate speech” more broadly? That is not, as many of you know, Parler’s approach. This is because the term “hate speech” is vague, and is generally held to encompass much speech that is protected by our First Amendment. We have all had a challenge, in the last couple weeks, determining what language, in which context, is “incitement” (“I love you”?) Imagine how subjective things will get when “hate speech” moderation becomes mandatory.
What has Facebook hoped to accomplish by encouraging this? I can only speculate that the company is trying to preserve both their data-mining practices and their rumored engagement-enhancing algorithms, upon which their monetization depends—and to hopefully keep it all under-the-hood, immune from discovery, via Section 230.
As for privacy, some say private companies don’t conduct “surveillance” when they enforce terms of service. Now I’ll remind you of the work I was doing before I joined Parler: promoting and deploying a novel solution to the problem of the “third-party doctrine.” The doctrine says that information a person shares with a “third-party”—such as a social media platform—is not protected by the Fourth Amendment, and therefore that government can obtain such information without a warrant.*
Now we can predict what’s to come: Section 230 will be amended as Zuckerberg suggests. Given the current capabilities of AI this will likely mean that all platforms will be required to scan ubiquitously for “hate speech,” “misinformation,” or who knows what else. The results of these scans will become ordinary business records of the platforms, obtainable by government without a warrant. No probable cause, no particularized suspicion—perhaps nothing more than a “consent order” could result in routine access to these records. Minority Report, anyone?
A few months ago, a colleague wondered whether we should engage in more activism. I said that just offering our product is plenty! At Parler, a crucial part of our mission is to collect only the bare minimum of user data, rejecting the business model of Big Tech as we know it today (with a few possible exceptions). In addition, given the total context—more of which I’ve come to understand only this week—we have been outspoken in calling for the repeal of Section 230. I hope more people will understand why we are seen as a threat to this industry—before it’s too late.
*This has been true only since the 1970’s, when the Supreme Court unjustifiably and without explanation transported the third-party doctrine from the context of information sharing in the course of criminal activity, to the context of information sharing within an ordinary business context.