Algorithm Opt-Out Laws Under Consideration in Australia

The Australian government is contemplating the implementation of laws that would require social media platforms to offer users an option to opt-out of algorithmic content recommendations. This initiative aims to mitigate the polarization that arises from content algorithms in social media applications. The consideration for such a measure follows the government’s recent efforts to address the issues surrounding social media usage among teenagers, which have, to date, shown minimal effectiveness in curbing the negative behaviors associated with these platforms.

According to a report from The Guardian, Australian social services minister Tanya Plibersek stated that the government is investigating the possibility of an algorithm opt-out. This exploration comes in response to tragic incidents, including a recent case of a teenage strangling death and an alarming rise in reports of sexual violence among adolescents.

In her comments to The Guardian, Plibersek highlighted a startling statistic: “If you set up a social media account as a young boy in Australia today, it takes an average of 23 minutes for you to be fed misogynistic, violent content online. That is a deliberate decision of the social media companies that is wrecking our young boys, giving them all of the wrong ideas about what healthy relationships look like.”

During an appearance on ABC’s Insiders program, Plibersek emphasized that the government is considering various strategies to tackle algorithmic amplification concerns. This includes exploring both opt-in and opt-out mechanisms that could allow users more control over the content they encounter on these platforms.

Tackling algorithmic amplification is crucial, as research suggests it may yield a more significant positive impact than merely banning teenagers from using social media applications.

Extensive academic research has examined the detrimental effects of algorithms, including:

• A 2016 study published in the Harvard Business Review revealed that high-arousal emotions, such as anger and fear, tend to provoke the most substantial social media engagement, particularly in terms of viral sharing.
• A 2023 study from the National Library of Medicine found that headlines containing negative language significantly increased engagement rates among social media users.
• A report analyzing Instagram interactions, published by Science Direct in 2021, identified that posts evoking negative emotions received up to 70% more likes compared to those eliciting positive feelings. accounts that generated higher engagement numbers had a considerably greater volume of likes and video views.
• An analysis by the Korean Advanced Institute of Science and Technology, published by Taylor and Francis Online in 2023, concluded that posts expressing anger reached a broader audience than those conveying anxiety. Users expressing anger were also more inclined to share emotional content.
• A 2012 study from Wharton Business School found that content that incites anger is more likely to be shared, with the intensity of anger correlating directly with the virality of that post.

Emotions like anger and fear drive algorithmic amplification because these systems rely on binary engagement metrics. More likes and comments influence algorithmic decisions, irrespective of the content’s nature, subsequently amplifying the visibility of these posts.

The evidence emphasizes that algorithms significantly contribute to triggering emotional responses, which can lead to increased division and distress. Given that a substantial portion of the population turns to social media for news and information, as highlighted by Pew Research, this situation poses a significant risk.

The primary challenge lies in implementing restrictions and compelling platforms to provide algorithm control options.

For instance, Meta has consistently promoted the advantages of algorithmic content matching regarding enhancing user engagement, which is intricately linked to its advertising revenue model.

Imposing restrictions on algorithms could disrupt its business structure, potentially leading to conflicts regarding trade regulations.

Since Australia is a relatively small market, it could serve as an effective testing ground for algorithm opt-outs. This would allow Meta and other regulators to assess the impacts of transitioning users to a purely chronological feed of posts from pages they have chosen to follow.

Meta has already piloted such an approach in the EU, where users can now pay a monthly fee to opt-out of algorithmic matching.

Despite this option being available, few users have opted to pay, rendering it a limited test case. A genuine test would involve the Australian government mandating an opt-in for algorithmic amplification, ensuring that all users are disconnected from the endless stream of recommended posts.

A comprehensive evaluation of this strategy could reveal its full potential, and it will be intriguing to see whether Australian regulators pursue this direction.