After threatening them with legal action, AlgorithmWatch researchers from Germany had to quit their Instagram algorithm study project.
AlgorithmWatch started a project in March 2020 to analyze Instagram’s newsfeed algorithm. Approximately 1,500 volunteers participated in the experiment over the course of 14 months. Installing an add-on that scraped their newsfeeds and sent the information to AlgorithmWatch. This is in order to discover how the business prioritized photos and videos on a timeline.
“Using their data, we were able to demonstrate that Instagram likely incentivized content producers to publish images that matched certain depictions of their bodies. Politicians might reach a bigger audience by avoiding language in their publications,” AlgorithmWatch says.
Both of these assertions are denied by Facebook. To be more specific, the first data point revealed that Instagram appeared to urge users to expose more skin than usual. Following their discovery, AlgorithmWatch sought out Facebook for comment. The social network firm first disregarded the researchers’ findings before informing them that it was “flawed in many respects.”
“While we couldn’t audit Instagram’s algorithm precisely. Our study is among the most sophisticated ever done on the platform,” AlgorithmWatch says.
The European Data Journalism Network and the Dutch organization SIDN sponsored the project. It is carried out in collaboration with Mediapart in France, NOS, Groene Amsterdammer, and Pointer in the Netherlands, and Süddeutsche Zeitung in Germany.
Facebook invited AlgorithmWatch to a meeting in May. The company told the group that they had violated the company’s Terms of Service. Also, if they did not “resolve” the issue on Facebook’s terms, Facebook would have to “move to a more formal engagement.”
AlgorithmWatch claims it chose to make this discussion public with Facebook. The social media giant deactivated the accounts of researchers at New York University’s Ad Observatory. That organization had created a browser add-on that gathered information regarding the platform’s advertising.
Facebook claims the researchers breached its terms of service. They collected illegal data from its network. According to the researchers, the business is trying to exercise control over any study that casts a bad light on it.
“This isn’t the first time Facebook has taken a strong stance against groups that want to empower people to be more independent in their usage of social media,” AlgorithmWatch adds. “It threatened Friendly, a smartphone software that allows users to choose how their newsfeed is sorted, in August 2020. It removed many applications from the Play Store in April 2021 that enabled users to access Facebook on their terms. There are very likely more instances of bullying that we are unaware of. We hope that by stepping out, other groups will share their stories.”
While AlgorithmWatch was forced to halt its research. They said that it is critical for organizations to shed light on Instagram’s algorithms and point to several instances where the business appears to be taking specific action against the emergence of certain types of information. This includes how both Colombian and Palestinian viewers noticed that content about ongoing protests in their country was removed from the platform.
“Large platforms have a significant, and generally unknown, impact on society. It ranges from identity formation to voting decisions. Only through more openness will we as a society be able to guarantee an evidence-based discussion about the role and effect of big platforms — a vital first step toward holding them accountable,” AlgorithmWatch adds.
I am extremely passionate about blogging, running websites, and creating content. I have managed to turn my passion into a profession, and blogging has managed to teach me a lot about technology and myself. I write blog posts, instruction manuals, news releases, and technical descriptions, and reviews for many websites such as norton.com/setup