A recent study reporting Facebook as a trustworthy source of news… has recently been proven inaccurate. The trusted news algorithms, on which the study was based, were dropped shortly after the study concluded However, new research is now suggesting the algorithm was altered to purposefully trick the study.
In 2023 research by Guess et al was carried out, declaring Facebook and Instagram as trustworthy sources of news that were not driving misinformation — ahead of the elections in 2020 when the data was first obtained.
However, since then, it’s been revealed that the Facebook algorithm’s preference for trusted news sources was quickly shelved, within months of the data being gathered.
There is now a concern that Facebook, who require advance knowledge of any study before it is undertaken, may have deliberately altered its algorithms, temporarily, to present a somewhat rosy picture to those investigating the platform.
Report Into the Report
A further report has since been published by researchers at the University of Massachusetts Amherst, in addition to Indiana University and University College Dublin. This report questions the findings of the 2023 study that suggested Facebook news feed algorithms presented a trustworthy source of information.
This more recent study explains that when the data was gathered, this was really a snapshot in time when the Facebook algorithm was temporarily favouring trusted news sources.
This algorithm was changed in March 2021, shortly after the original study had been concluded in December 2020. The findings were published, and Facebook was deemed to be perfectly legitimate with no risk of spreading misinformation.
Whilst the report was accurate for that specific time period, it certainly did not reflect the nature of Facebook during the years that followed, or its characteristics in general.
So why exactly would the team at Facebook want to change its algorithm to stop preferring trusted news sources when deciding what to show to the user?
Is Fake News More Profitable Than the Truth?
To understand the factors that helped Facebook reach its decision to tweak its trusted news algorithm, it’s important to understand how lucrative fake news can be for organisations, and in particular, social media companies.
Whilst they may not generate the fake stories themselves, they are certainly not averse to the increased revenue it brings them. But why is fake news more profitable?
Understanding the Mechanics Behind Fake News Profitability
To understand why a completely fabricated story can generate more income than the truthful version of events, we need to look at how this works within the machine of a social media platform such as Facebook.
Take, for example, the classic “eating the dogs, eating the cats” quote from the recent American election campaign.
During the presidential debate, President Trump declared that there were big problems in places like Springfield. He had heard on a fringe news channel that a woman had reported her cat was missing and was likely being eaten by a Haitian gentleman.
In reality, the cat was not eaten at all and was in fact later found in the basement alive and well. This was the correct version of events, although this was drowned out by a more shocking narrative.
Social media is fuelled by content that is the most entertaining and interesting to its users - this does not always coincide with factual correctness or accuracy.
A cat being reported missing and later being found in a basement is not especially intriguing and will therefore not get a lot of traction. It would not prompt people to share this with their connections, and in the process, it would fail to create a viral trend.
However, the much more controversial story about cats being eaten was amplified across the globe, even by people who did not necessarily believe this version of events and ridiculed the idea. There were memes and even songs produced by various artists.
These posts and the virality surrounding them can be great for social media companies themselves. Put simply, popular posts are profitable posts.
When a video on YouTube, for example, gets millions of views there is a swathe of commercial activity generated surrounding it: sponsorship and affiliate deals are struck, adverts are placed on the content, and the social media platform will normally take a share of the revenue generated for the advertiser via the platform.
In this way, fake news can be a big money maker for any social media platform prepared to allow inaccuracy to run rampant on their site.
The same mechanism works for any fake story—whether it is a post about Hillary Clinton being involved with alien terrorists, or somebody explaining that the moon landings were, in fact, a conspiracy. If it sounds interesting and juicy, there’s a good chance it will be amplified across social media, regardless of the validity of the facts and methods of reporting.
How did the Algorithm Change?
During the 2020 US elections, Facebook wanted to ensure that misinformation could not be spread excessively throughout the platform and refined their algorithms so that when a user opened up their news feed the system would check whether content sources were from a trusted news outlet.
This could for example include the BBC, CNN, New Scientist, or the National Geographic, These validated sources of news would receive preferential treatment within the algorithm, rising to the top of the feed.
This would ultimately shape what users would see and the version of events they would come to understand.
This was great for accuracy and the democratic spread of evidence however, there was one issue—it was not quite as good for generating revenue.
Facebook looked at the figures and found that, since the adaption of it’s algorithms, engagement was on the decline.
Just like in a soap opera, the truth of the matter is that drama sells.
After looking at this drop in revenue, Facebook quickly adapted the algorithm again in March 2021, only 3 months after the report was concluded.
Fake news was back on the menu.
Social Media Impact on Democracy
The study carried out from September 2020–December 2020 only saw a snapshot of Facebook’s news feed algorithm, which had been temporarily altered.
Some have argued that, as the studies need to be “preregistered” with Meta, this allows them to adjust their algorithms to a more favourable setting before being called into question.
It seems more likely however that Facebook altered the news feed algorithm for the stated purpose: to avoid the spread of misinformation during the 2020 elections.
What this does mean is that the original findings of the study are not representative of Facebook in general
How Could the Study be Improved?
In order to improve the accuracy of the study, a more detailed investigation would need to be carried out, taking data at different intervals and comparing the results.
For example, if you studied the trustworthiness of the news feed algorithm in 2020, 2022, and 2024, you could start to develop a better picture of how the platform has changed as algorithms have been updated, spotting any trends that could suggest a move towards less trustworthy news being more prolific on the site.
Which Conclusions Can Be Drawn?
The more recent study from the University of Massachusetts Amherst, Indiana University, and the University College Dublin, suggests that debunking the findings from the previous study by Guess et al shows there are challenges in studying social media platforms, especially when they have to be forewarned, which allows for alterations to the system.
You could go a step further and say that all the findings from the original report are invalid and social media such as Facebook, Twitter (X), Instagram, TikTok etc, do indeed drive misinformation and polarisation of views.
This is not necessarily through some conscious or malicious act on the part of developers, but simply due to the way the platforms operate, with more popular and engaging content rising to the top, allowing external influences to take hold and manipulate the facts.
However, there is still further work to be done in this field in order to ascertain the true impact of social media on democracy.
Get the Latest Insights on Technology
For all the latest news on technology and how it’s set to change in the future, check out our news section or get in touch with our team directly to learn detailed insights into the industry you are operating in.
At Lyon Tech, we provide a range of technological solutions and high-end services for any business looking to stay competitive in an ever-evolving marketplace.
