The Record

Reduction / Borderline Content / Shadowbanning

Reducing the visibility of risky, misleading, or salacious content is becoming a commonplace and large-scale part of platform governance. Using machine learning classifiers, platforms identify content that is misleading enough, risky enough, problematic enough to warrant reducing its visibility by demoting or excluding it from the algorithmic rankings and recommendations.

On “Trusted” Flaggers

This essay unpacks the practices of trusted flagging. We first discuss self-regulatory flagging partnerships on several major platforms. We then review various forms of government involvement and regulation, focusing especially on the EU context, where law-making on this issue is especially prevalent. On this basis, we conceptualize different variants of trusted flagging, in terms of their legal construction, position in the content moderation process and the nature of their flagging privileges.

Forgetful Advertising: Imagining a More Responsible Digital Ad System

As Silicon Valley giants sketch their preferred future for digital advertising, an infrastructure with significant implications for life online and offline, there are startlingly few alternatives to their vision. In response, we propose “forgetful advertising”, a vision for responsible digital advertising structured around a single design choice: avoiding the storage of behavioral data.

Section 230 and the International Law of Facebook

Despite its local origins, Section 230 serves as a key architectural norm of the global internet: internet service providers are not typically responsible for the speech of their users. Section 230 underpins what we might describe as the International Law of Facebook— the global community guidelines written and enforced by internet platforms, largely allowing these platforms to regulate the speech on their platforms. Reviewing Section 230 cases involving foreign events, foreign parties, or foreign law, the essay reveals how Section 230 made the U.S.

Sonic Privacy

Advances in data collection and processing have facilitated ultra- and infra-sonic machine-listening and learning. This requires the recognition of sonic privacy, protection for our “sonic data:” those representations or observations that define the characteristics of sound and its cognitive and emotive forces. This right would protect (non)participation in the public sphere.