-- LauraBane - 20 Oct 2024

Killing (The Algorithm) In The Name (Of Liberation)

Introduction

Imagine that you are a twelve-year-old girl who has access to an Internet-capable device and is not subject to Panopticon-level supervision by her parents. Feeling the weight of peer pressure, you download social media app Instagram. Over the next several days, you will be inundated with posts which will permanently alter your self-conception and give rise to myriad new insecurities. In your pre-social media life, attractiveness was Boolean—people were either fat or thin, pretty or ugly. Now, your mind stirs with thoughts of buccal fat, thigh gaps, strawberry legs, upper lip-to-gums ratios, and 'P-' versus 'D-' shaped silhouettes. You stop eating, beg your parents to buy you retinol cream and let you save up for cosmetic surgery, and stare at yourself in the mirror until you no longer recognize your reflection. In any other context, this sort of disruption and anguish would be deemed an impermissible psychological operation, the stuff of conspiracy theories like MK Ultra. Yet, because of the misconceptions surrounding technology-fueled socialization and the ultra-libertarian legal landscape which governs it, many view the rise in eating disorders and suicidal ideation among children and young adults as an inevitable price to pay for the “convenience” of online social networking.

The Relevant Legal Landscape

To understand this psychologically violent phenomenon and identify solutions, one must first understand the laws which have allowed it to flourish. Because social media posts are undoubtedly forms of expression,_1_ they are subject to First Amendment jurisprudence. This means that despite social media platforms’ status as private companies and their consequent right to restrict constitutionally protected speech and expression such as racial slurs or nudity, the government cannot force them to regulate undesirable, but legal, speech._2_ And, even in the case of unprotected speech and expression, such as child sexual abuse material (“CSAM”) or imminent threats of violence, social media companies are largely insulated from liability as a result of §230(c) of the Communications Decency Act of 1996._3_ §230(c)(1) prevents “interactive computer service” providers from being deemed the “publisher or speaker of any information provided by another information content provider.” Id. §230(c)(2) protects “operators of interactive computer services” who remove, in good faith, material that is “obscene, lewd, . . . excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally protected.” Id. One key factor in applying §230(c) immunity to social media platforms is whether said platform engaged in the speech itself, or whether it simply provided a space for others to speak. For example, the Third Circuit recently permitted the lawsuit of a mother whose child died as a result of toxic messaging on the social media app TikTok? to proceed, stating that because TikTok? curates its algorithm before presenting it to users, it is a publisher or speaker, therefore rendering §230 inapplicable._4_ Another important element of §230(c) analysis is deciding what constitutes the “good faith” removal of objectionable content. Despite harrowing testimony at a congressional hearing earlier this year that Mark Zuckerberg’s Meta apps “pushe[d] pro-eating disorder content to vulnerable populations”_5_ and gave users a lukewarm warning that certain search terms may reveal CSAM but allowed users to “[s]ee results anyway,”_6_ Zuckerberg has not been held sufficiently accountable._7_ Zuckerberg has paid lip service to curbing his platforms’ negative effects on minors, such as requiring all Facebook users to certify that they are at least thirteen years old, but whistleblower reports have shown that children can simply lie about their age when creating an account. Id. And, although Facebook “could employ the mechanisms it uses to analyze other types of audiences on the platform” and eliminate underage users, it “just chooses not to do so.” Id.

Potential Solutions

In 2020, the Department of Justice issued recommendations to Congress to modify §230._8_ These recommendations included (i) “incentivizing platforms to deal with illicit content,” primarily by publicly shaming companies that “solicit illicit activity,” (ii) removing protections for sites known to “host child abuse, terrorism, and cyber-stalking,” including any site which has “been notified by courts of” the existence of such material, (iii) removing protections from civil lawsuits initiated by the federal government, and (iv) more clearly defining “good faith” in the context of the statute. Id. For the immediate future—assuming that social media platforms will continue to be run by billionaires and attract millions of users across the country—I propose two alternatives, partially because I believe that they will effectively target speech which does not rise to the same level of legal actionability as child abuse or terrorism but is nonetheless undesirable. First, minors’ use of social media should be regulated in a manner similar to drinking—there should be a legally enforceable minimum age requirement. Second, social media platforms should be required to ask all users for photo identification in order to verify users’ identities. This will not only ensure that underage people are not using the platforms, but also disincentivize the posting or sharing of illegal material, such as CSAM. However, I am deeply uncomfortable with social media platforms having access to information such as one’s home address or social security number. For that reason, I propose a longer term solution: the creation of decentralized online networks by which people can communicate with one another and edit each other’s posts far more freely than they can on current social media platforms. At first, it seems counterintuitive—didn’t I say that a lack of regulation was the problem? Yet, here, regulation exists—it just exists in the hands of many. Without the overarching power imbalance between a single owner who wants to make money by partnering with producers of consumer goods and millions of disenfranchised users, people would be able to self-select to act as moderators, and the perverted ‘algorithm’ would cease to exist. Instead, people would have to make a conscious choice to seek out online fora for their interests and could remove content which they find disturbing. And, because everyone would be speaking self-representatively, rather than amplifying others’ speech (as Zuckerberg does under the current social media model), §230 would be rendered obsolete, thereby facilitating increased accountability.

You use far too much space summarizing legal material that can be addressed with one link. But you do not explain at all the central legal hypothesis of the draft: that regulation is necessary. So long as you have a working legislative majority in both houses sufficiently hostile to the platforms and a president unwilling to veto the legislation (conditions which might be expected to exist, I suppose, were it not for Elon Musk serving as de facto VP), why not just repeal 230 altogether? No one except platform owners thinks they deserve an immense subsidy through immunity from ordinary rules of liability. Those rules, if back in force, would pervasively alter their behavior in the desired direction, overnight.

So the primary problem is that Musk helped to buy the presidency for Donald Trump to ensure that 230 immunity remains in place. Wouldn't it be useful to explain that fact, to contest it if you doubt it, and in either event to reason directly about he consequences rather than speculate counter-factually?

Sources

Why aren't these just links anchored to the relevant text? The Web allows us to maske it possible for the reader to reach the source with a click, and the wiki makes all linking easy. So why the cumbersome three-step process to do what comes naturally?

1. Elonis v. U.S., 575 U.S. 723 (2015). 2. https://www.freedomforum.org/free-speech-on-social-media/ 3. https://en.wikipedia.org/wiki/Section_230 4. https://www.wsj.com/us-news/law/appeals-court-raises-questions-over-section-230-law-giving-social-media-companies-legal-immunity-af4c1e6c 5. https://www.klobuchar.senate.gov/public/index.cfm/2021/10/at-commerce-committee-hearing-with-facebook-whistleblower-klobuchar-highlights-how-facebook-algorithm-promotes-eating-disorders-among-young-users 6. https://www.judiciary.senate.gov/press/releases/judiciary-commerce-committee-leaders-want-answers-from-meta-on-pedophile-network-story 7. https://time.com/6104070/facebook-whistleblower-congressional-hearing-takeaways/ 8. See footnote 3.