From Platforms to Campuses

Content moderation moves from online to university campuses; holding platforms accountable for asymmetrical reciprocity; and, an affiliate wins an award!

Content moderation on campuses

“Trust and safety professionals at social media companies have spent the past 20 years working through some of the hardest questions about how to police speech, and universities may be able to take lessons away from their experiences.” Matt Perrault, director at the UNC Center on Technology Policy, and Daniel Kreiss write for Tech Policy Press on content moderation on college campuses. The debates surrounding campus speech policies have intensified, notably after instances where university leaders were’t able to explain previous decisions in a congressional hearing, causing concerns about the clarity of campus speech policies. Recent events at UNC Chapel Hill, where protests led to arrests, have spotlighted the ongoing challenges faced by universities in moderating speech effectively. This situation reveals a persistent uncertainty among higher education institutions about establishing a clear and functional framework for speech on campus.

Perrault and Kreiss then bring attention to the parallels between platforms and universities when it comes to content moderation, noting, "Tech companies, like universities today, have been called on to remove legal speech that some people believe poses a safety threat to the community or some groups within it." Social media companies have navigated complex issues around speech and safety, adapting their policies over time to balance these concerns with the need for free expression. Due to this precedent, Perrault and Kreiss suggest college leadership look to platforms for guidance: “Learning from these approaches might help universities to better navigate the speech challenges they are facing.”

In particular, universities might enhance their handling of campus speech by incorporating four core principles demonstrated by social media companies:

• Establish clear policies. “Universities should evaluate their policies through the lens of whether speech will result in actual harm to the learning experience, not simply if it is offensive or upsetting.”

• Apply thoughtful remedies. “A commitment to education means that universities should avoid binary approaches to punishment that are too lenient (ignoring problematic conduct) or too heavy-handed (arrests and expulsion)… For certain categories of sensitive speech, universities could permit it only in certain locations (essentially a “time, place, or manner” restriction) so it does not disrupt their educational missions.”

• Maintain transparency. “Universities could publish statistics on their enforcement of campus speech policies and provide detailed rationales for speech-related decisions, including reports that retroactively analyze campus controversies which stakeholders can then convene around.”

• Engage in ongoing research. “Universities should conduct research to better understand campus speech harms and to evaluate the efficacy of their speech policies. They should conduct impact assessments to understand the costs and benefits of their speech policies for university communities. And they should solicit feedback from a broad range of diverse stakeholders in the field, and then use these learnings to improve their practices over time.”

By adopting these elements, universities could better manage speech-related challenges, ensuring that their policies not only respect free expression but also safeguard the community from speech that could harm the educational environment.

Asymmetrical reciprocity

Jacob Smith, CITAP GRA, along with co-authors Aaron Shapiro, Courtlyn Pippert, and Zari Taylor, explore the intricate dynamics of power and reciprocity between digital platforms and their users through the lens of "asymmetrical reciprocity," in “Patrons of commerce: Asymmetrical reciprocity and moral economics of platform power”. The authors posit that platforms often position themselves as benevolent patrons, creating a perceived imbalance where they appear to give more than they take. This perception helps platforms legitimize their authority and control, but it is a fragile legitimacy that users frequently challenge. The authors argue that platforms utilize this patronage to mask the true nature of their interactions with users, where they often extract more value than they contribute.

Case Study #1: Twitch

Changes in subscription revenue policies sparked significant backlash on Twitch; the platform's decision to alter the split of subscription revenues from a favorable 70/30% to a less generous 50/50% for top earners was met with outcry from content creators. These creators felt that Twitch, by rolling back previously negotiated terms, had violated a tacit agreement of mutual benefit and support. This vignette illustrates the precarious balance platforms must maintain between their profitability and their public image as supportive of their user communities.

Case Study #2: Amazon Kindle

Authors utilizing Amazon’s Kindle Direct Publishing faced challenges with scammers manipulating the platform's payment algorithms. The platform's initial reluctance to address the scamming issues effectively showed a discrepancy between the support Amazon claimed to offer to authors and the reality of their policies. This situation highlighted the tension between Amazon's role as a market leader in digital publishing and its responsibilities toward the creators who populate its platform with content.

Case Study #3: Deplatforming adult content creators and sex workers

Sex workers and adult content creators, who often face harsher content moderation compared to other users, are directly impacted by shadowbans and account suspensions on platforms. This segment of platform users regularly contends with arbitrary enforcement of policies, which disproportionately affects their ability to earn a livelihood. The response from the platforms often lacks transparency and fails to address the specific needs and challenges faced by these creators, revealing a significant imbalance in how platforms manage relationships with different user groups.

How should platforms address these discrepancies in power?

The authors argue that to rectify the asymmetries in power and reciprocity, platforms need to invest in what the authors term "legitimacy costs." Platforms can better align their operations with the interests of their diverse user communities, thereby stabilizing their legitimacy and fostering a more equitable digital ecosystem.

Publications and Appearances

Deen Freelon coauthored the study, "The effects of Facebook and Instagram on the 2020 election: A deactivation experiment"; the researchers investigated the impact of social media on political engagement and misinformation by having a sample of Facebook and Instagram users deactivate their accounts for six weeks prior to the 2020 U.S. Presidential election. Their findings indicated that while deactivation led to a decrease in online political participation, it had no significant effects on knowledge, polarization, or the perceived legitimacy of the election. The study suggested that while deactivating Facebook decreased exposure to general news and misinformation, it did not significantly affect users’ voting behavior or their overall perceptions of electoral legitimacy. This experiment provided nuanced insights into the complex role of social media in shaping political outcomes, highlighting both the limits and the potential influence of platforms like Facebook and Instagram during election periods.

Francesca Tripodi joined NYU Center for Social Media and Politics’s “The Future of Search Engines in the Age of AI” on May 15^th. Francesca, Zoe Darme from Google, Mia Sato from the Verge, and Zeve Sanderson from CSMaP discussed the role of AI in shaping access to information and how the implementation of these technologies affect trust, misinformation, and public knowledge. Francesca discussed data voids and the relationship between information databases and AI responses, noting:

“Nefarious actors will construct key words very explicitly to connect to those deep stories. Then they spread these narratives in a non-necessarily online environment, right? So talk radio is still a primary way where people get information in the United States, in-person rallies and events. Then there's this distrust of mainstream media and this huge emphasis to do your own research. But these research rabbit holes are often highly structured in a way that people don't necessarily understand, but when they search for these highly curated key words on these concepts, the information returned to them is not particularly great.

To me, the most pressing concern in an AI-enabled search environment are actually data voids, because it's still unclear how AI enabled search is being trained. We know there's a very high link between Wikipedia content and Google Google results, but we don't really know how this is mapping into AI-enabled search.”

Affiliate Highlights

Rohan Grover, doctoral candidate at the Annenberg Schol for Communication and Journalism at the University of Southern California, was awarded the first-place winner of the annual James R. Cleary Prize from the UNC Center for media Law and Policy.

His article, “Contingent Connectivity: Internet Shutdowns and the Infrastructural Precarity of Digital Citizenship”, “evaluates the implications for citizenship itself by bringing together scholarship on digital governance, science and technology studies (STS) approaches to Internet governance, and postcolonial and decolonial theory. Notably, this article raises the stakes for critical analysis of how authoritarian states approach Internet policy to bridge digital divides—and for evaluating quality and contingency of connectivity experienced by marginalized and peripheral communities.” Congrats, Rohan! 👏👏👏

Coming soon

May 21^st @ 6:30pm: Tressie McMillan Cottom is participating in a Center for Brooklyn History talk with Natalie Foster on Foster’s “The Guarantee: The Fight for America’s Next Economy.” It will be in person in Brooklyn- more info here.