Summary: Integrity Institute’s Amicus Brief on Gonzalez v. Google

Dec 9, 2022

In support of neither party, the amicus brief explains major uses of algorithms in large tech platforms and urges the Court to decide narrowly based on technical specificity and nuance.

• Our announcement is here

• Full text of our amicus brief is here

Overview

• The Integrity Institute is a think tank powered by a community of integrity workers, many of whom have helped build recommender systems (colloquially “algorithms”) used in large tech platforms. Our amicus brief, in support of neither party, offers this community’s independent, professional expertise to the Court.

• The case of Gonzalez v. Google concerns whether or not recommender systems are covered by liability exemptions under Section 230 of the Telecommunications Act of 1996 for Internet service providers. Backed by the Integrity Institute’s member expertise, our amicus brief explains to the Court how tech platforms actually use algorithms.

• Algorithms are core components for tech platforms to function, and platforms use different types of algorithms depending on contexts (such as content recommendation, content moderation, or advertising). Well-designed algorithms can in fact enable a functional and enjoyable experience when using tech platforms.

• Given the specificity of different algorithms, this amicus brief urges the Court to focus on the specific facts and algorithms in this case to decide on an individual, narrow basis.

Three types of algorithms explained

This amicus brief explains three types of algorithms commonly used in large tech platforms, as well as their respective harms and benefits, to illustrate to the Court that technical specificity and nuance are of utmost importance in cases involving algorithms and the question of liability.

Algorithms for content recommendations

• Platforms may use algorithms to recommend content in ways that personalize recommendations for individual users based on their past behaviors (as well as inferred characteristics) and optimize expected value to the company by maximizing individual users’ expected engagement with recommended content.

• When platforms use algorithms to maximize engagement, they cannot fully prevent harmful third-party contents from being recommended to users if those users have consumed similar contents in the past.

• Algorithms used for content recommendations can benefit user experience in scenarios such as recommending better matches on dating apps or autofilling URLs on web browsers.

Algorithms for content moderation and safety

• Platforms use algorithms to prevent and reduce harms by semi-automating the process of flagging, removing, and re-ranking third-party contents likely to violate platform policies or laws. When this process is performed at scale, the algorithms cannot perform perfectly and are continuously optimized to balance between precision and accuracy.

• If a platform prioritizes accuracy over precision in using algorithms for content moderation, its process would have a high false positive rate. Most large platforms therefore choose to prioritize precision over accuracy, which allows most users to post contents but can sometimes lead to extensive harm when false negatives are shared widely.

Algorithms for advertising and commerce

• Platforms use algorithms to serve targeted ads to individuals through “retargeting,” which relies on expressed and inferred information about those individuals that the platforms had already compiled.

• Algorithms that are used in techniques like “retargeting” primarily benefit companies, and this encourages companies to collect more and more data about users.