The Responsible AI Bulletin #23: Authorship identification, trust in AI, and flagging social media posts.

Welcome to this edition of The?Responsible AI Bulletin, a weekly agglomeration of?research developments?in the field from around the Internet that caught my attention - a few morsels to dazzle in your next discussion on AI, its ethical implications, and what it means for?our future.

For those looking for more detailed investigations into research and reporting in the field of Responsible AI, I recommend subscribing to the AI Ethics Brief, published by my team at the Montreal AI Ethics Institute, an international non-profit research institute with a mission to democratize AI ethics literacy.

Defending Against Authorship Identification Attacks

Suppose you just discovered that your company is engaging in unethical activities. Driven by a sense of justice, you decide to blow the whistle. To avoid retaliation, you choose to post a few lines on social media using a throw-away email and a camouflaged IP address. Before hitting ‘enter,’ a thought strikes you: does this precaution truly guarantee anonymity?

The answer is NO. Studies have shown that a person’s writing style can reveal their identity.?

Based on the most frequent word distribution, one basic machine learning method correctly predicts the author 70% of the time from a pool of 40 candidates. The probability of evading linguistic forensics is slim given a company’s access to an employee’s past emails and reports.?

In this digital era, every text, whether a tweet, blog post or research paper, can potentially be used to trace its author’s subsequent writings, a task known as authorship identification. Abuse of authorship identification raises significant privacy concerns, particularly for whistleblowers, journalists, activists, and individuals living under oppressive regimes.

Continue reading here.

Two Decades of Empirical Research on Trust in AI: A Bibliometric Analysis and HCI Research Agenda

Would you trust AI to pick your next vacation spot, give you health advice, or drive your car? Researchers have been exploring questions like these to better understand how or to what extent people trust different AI tools. To contribute to a better understanding of the research landscape and determine which research trends hinder or foster progress in our understanding of this complex concept, the authors conducted a comprehensive bibliometric analysis of two decades of empirical research on trust in AI across various disciplines, uncovering publication patterns, and describing the underlying knowledge structure of the field. They highlight and discuss several trends concerning (a) the rapidly evolving and increasingly heterogeneous research dynamic and the main research themes, (b) the foundational works and research domains that empirical research leverages, and (c) the predominant exploratory nature of empirical research on trust in AI. In light of these trends, the authors outline a research agenda facilitating a shift from exploration toward developing contextualized theoretical frameworks. They argue that such a shift is crucial to cultivate an in-depth understanding of trust in AI, one that can serve as a foundation for practitioners and inform the design of safe and trusted AI.

Continue reading here.

Cleaning Up the Streets: Understanding Motivations, Mental Models, and Concerns of Users Flagging Social Media Posts

NOTE: see figure references in the detailed post.

How do social media platform users provide feedback to platforms when they view harmful content? Despite the ever-changing nature of online platforms, the mechanisms by which users can report inappropriate content have remained relatively limited and unchanged over the past three decades. Flagging is the primary action users can take to report content they believe violates platform guidelines. Flags are often presented as icons users can click to report or “flag” harmful content. Most social media platforms offer guidelines on what constitutes harmful content, often termed “community guidelines.” These guidelines cover many of the well-known categories of harm, such as sexual content, violent content, harassment, misinformation, spam, and more. However, it is often unclear to users when to engage in flagging or what occurs when they decide to flag content. For example, Figures (a) through (c) showcase the flagging process on YouTube, during which the only feedback users receive for flagging content is a “Thanks for reporting” message. Note that while we call these mechanisms “flags” in our research, many platforms use other terms, such as “report,” to refer to the same mechanisms.

In contrast, Figures (d) through (e) show how Facebook prompts users to take further action when they flag a post, such as asking if they want to block the account that posted the harmful content or hide all of their posts. In this way, Facebook’s more sophisticated flagging process allows users to incorporate other moderation tools, such as blocking and hiding content, to further protect themselves from harmful content.?

Figures (f) and (g) show Facebook’s Support Inbox, an interface that allows users to track what happens to the content they report. In contrast to YouTube, which provides only a thank you message, Facebook’s model offers users a separate interface dedicated to tracking the status of their flags. These two examples show that even among major social media platforms such as Facebook and YouTube, there are discrepancies in how platforms approach flagging. This, in turn, may affect how platform users choose to engage in flagging.

To understand flagging from platform users’ perspectives, we interviewed 22 active social media users who recently reported content to understand their perspectives on flagging procedures. These factors motivate or demotivate them to report inappropriate posts and their emotional, cognitive, and privacy concerns regarding flagging. Our analysis of this interview data shows that a belief in generalized reciprocity motivates flag submissions. Still, deficiencies in procedural transparency create gaps in users’ mental models of how platforms process flags. We highlight how flags raise questions about distributing labor and responsibility between platforms and users for addressing online harm. We recommend innovations in the flagging design space that provide greater transparency about what happens to flagged content and how platforms can meet users’ privacy and security expectations.

Continue reading here.

Comment and let me know?what you liked and if you have any recommendations on what I should read and cover next week. You can learn more about my work here. See you soon!