Ethics of Disinformation Copy

Moral Lenses Analysis

When it comes to the ethics around disinformation, we can apply a few main moral lenses to help us analyse complex situations. See the examples below in the context of a case study:

Case Study: Palestinian Content on Social Media

Social media platforms (e.g., Facebook, YouTube, and Twitter / X) combine AI and human efforts for moderation of their platforms.

Facebook’s AI uses strategies such as “remove, reduce, and inform” to manage content that breaches its terms of service. Techniques like Neuro-Linguistic Programming (NLP) and Panoptic FPN are utilized to identify and process content related to nudity, violence, child pornography, and terrorism, which are automatically removed.

Meanwhile, content categorized as misinformation might be reduced in visibility, and content deemed violent or sensitive is flagged to inform viewers.

Despite these measures aiming to comply with human rights standards, the use of AI in content moderation has led to errors and controversy, particularly with the wrongful removal of Palestinian content, highlighting concerns about digital discrimination and the accuracy of AI systems.

In the MENA region, a notable case study involves the alleged censorship and removal of Palestinian content by major social media platforms using AI-driven moderation tools.

Reports have surfaced suggesting that AI algorithms disproportionately flag and remove Palestinian posts, especially during heightened periods of conflict. Activists and digital rights organizations argue that these AI systems, trained potentially with biases that do not adequately distinguish between legitimate political expression and prohibited content, result in unfair censorship.

This has sparked significant debate about the neutrality and effectiveness of AI in content moderation, calling for more transparent and accountable moderation practices that respect users’ rights to free expression while effectively identifying and addressing actual policy violations.

People illustrations by Storyset

Moral Lenses Applied to the Case Study

The Utilitarian tradition

The Utilitarian tradition evaluates actions based on their consequences, aiming to maximize overall welfare. For social media platforms, this means considering the balance between harms and benefits when deciding on content moderation levels. In our case study, platforms leaned towards stricter moderation due to concerns over risks like incitement and misinformation, though the underlying assumptions behind these decisions are open to questioning.

The Common Good approach

The Common Good approach focuses on the overall well-being of the community, highlighting values such as respect and compassion for all individuals, particularly those who are marginalized. When it comes to moderating social media discussions on the Gaza conflict, the goal is to support the oppressed and maintain communal unity. However, the strict enforcement of AI moderation measures may inadvertently silence Palestinian voices and raise concerns about whether it is truly beneficial for the community as a whole. Are there any individuals who are being left out of this community-driven prosperity due to moderation decisions?

The Rights approach

The Rights Approach emphasizes the fundamental dignity and freedoms of individuals that should always be respected. It can be challenging to navigate the rights of different groups, like social media platforms, journalists, civilians in Gaza, and users interested in news coverage of the conflict. Deciding which rights take precedence involves weighing potential conflicts and ethical considerations.

The Virtue approach

The Virtue Approach focuses on promoting admirable character traits, like honesty, when making decisions. Transparency is crucial for social media platforms dealing with content moderation challenges. Honesty, in this case, involves openly communicating moderation procedures and rules to users, and guaranteeing clarity and responsibility in decision-making. So, for Facebook to demonstrate virtue in content moderation, they need to prioritize transparency, giving users the opportunity to comprehend and acknowledge the reasoning behind their moderation choices.a

Justice

Justice is extremely important in assessing decisions made by social media platforms about what content to moderate. It means treating everyone fairly and explaining any differences in treatment. Considering justice also involves thinking about whether moderation policies take into account all the people who are affected. If they don’t, there may be questions about fairness and the standards used to make decisions.

The practice of virtue ethics involves a continuous balancing act. There is no comprehensive handbook for navigating every moral dilemma that calls for virtuous behavior. Applying this theory is a unique process for each situation, with no two cases being the same. While honesty and justice may guide social media companies in content moderation, patience and temperance could be more fitting in a different scenario. How can we determine which virtues to prioritize in a specific case? And how should we handle conflicting virtues that offer conflicting solutions?

(adapted with the permission of the Markkula Center for Applied Ethics, www.scu.edu/ethics, https://www.scu.edu/ethics/focus-areas/journalism-and-media-ethics/resources/the-ethics-of-social-media-decision-making-on-handling-the-new-york-post-october-surprise/)

GET BAD NEWS

https://getbadnews.com

Time to play the game! How good are you at spreading disinformation?