MANILA, Philippines — Social media platforms' overemphasis on weeding out false content to counter disinformation failed to combat the coordinated spread of false and twisted narratives in favor of President Ferdinand Marcos Jr. during the 2022 elections, a new study on disinformation has found.
A new report by digital rights organization DigitalReach found that the online disinformation that helped catapult Marcos to the presidency — which had been "years in the making" —- was widely disseminated by a network of trolls and credentialed figures that platforms could not stop from reposting dubious content that had already been taken down.
Related Stories
Even as they partnered with the Commission on Elections, social media platforms still "overlooked the dynamics" of how bad actors "went beyond efforts to misinform and disinform Filipino voters … to sow distrust and sap confidence in the electoral system and voting process at large," according to the study.
Gaps in social media platforms' efforts to stem the spread of disinformation and to align with how disinformation campaigns are done through influence operations "played a crucial role in election manipulation" in 2022, the authors of the study wrote.
"Mitigation efforts by social media platforms during the election were superficial. They failed to deeply consider the dynamics of the situation—namely how platforms could be weaponized during the election," the study stated.
Takedowns 'ineffective'
The study, titled "Operating the Propaganda Machine: Social Media Accountability in the 2022 Philippine Election," documented how pro-Marcos accounts took advantage of the ease with which content can be reposted across platforms as well as social media giants' limited policies on political ads.
Social media platforms' had announced prior to the elections that they would remove content deemed disinformation and misinformation. But "the intensity of information disorders and their widespread dissemination via coordinated networks" showed this approach was "largely ineffective," the study said.
"Typically, information disorders in the Philippines are systematically disseminated via networks where different accounts amplify each other's messages, and where the posting and reposting of content across platforms by different accounts leads to greater visibility and a longer shelf life," the study stated.
"As a result, simply removing content does not work, as content removed from one account or platform can resurface shortly afterwards on another," the report added.
The proliferation of short-form videos and memes that often repurposed a false or misleading narrative also contributed to the difficulty of moderating the content being disseminated across platforms.
The same claims had to be repeatedly fact-checked as they constantly cropped up on various platforms and in different formats.
Content in short-form video format is also "hard to moderate because it can involve nonverbal languages and is often broadcast live," the study noted.
"While social media platforms made efforts to handle information disorders in collaboration with fact-checking entities and initiatives, the battle against malicious content was lost. It appeared in such high volumes that it was able to dominate the platforms," the report added.
The study added that fact-checkers and researchers "often found that content was not removed even after it was flagged."
"For example, YouTube was given a list of videos and channels flagged by researchers for spreading historical distortions related to the Marcos family. YouTube executives never got back to the researchers, despite giving assurances that they would," the report said.
Meanwhile, the strategic dissemination of "borderline content" — content that nearly violates social media guidelines but does not contain outright false information — helped keep disinformation alive.
The report cited as an example of borderline content the online narratives painting Marcos as a "loving family man" and other videos that appealed to audience's emotions rather than spewing false claims.
"How to approach this kind of content is a matter of debate, as removing it can be considered as excessive or harmful to freedom of expression," the study said.
"This type of content is not obviously harmful and does not qualify as disinformation and misinformation or hate speech, but, as part of the family's rebranding, has a hidden message," the study added.
Policies on political ads outdated
Social media giants' policies regarding political ads also "did not have a significant impact on the tackling of information disorders in the election," according to the study.
Meta's political ads policy of declaring the person behind the advertisements — which it said had helped voters become better informed — was "not enough to keep the platform free of toxic information," the authors of the study noted.
"Disinformation and misinformation were still being advertised on the platform, and Facebook still received money from it," the study said.
The report added: "The situation is especially concerning as the platform appears to have profited from the dissemination of toxic content."
Other social media platforms that decided to outright ban political ads, such as TikTok, Twitter and Youtube, also failed to prevent the coordinated spread of disinformation related to the elections, many of which are posted or uploaded by non-candidates.
"Political figures have been shown to be the clients of companies that handle political campaigns, including by designing campaign messages, executing strategies for message dissemination, and hiring influencers and anonymous bloggers," the study said.
While the study noted that social media platforms are not the sole reason behind the disinformation-laden discourse that helped Marcos win the elections, "it is very likely that Marcos Jr. would not have won were it not for his weaponization of social media."
The authors said that the outcome could have been prevented had social media platforms been more transparent in their efforts to prevent the spread of disinformation.
The social media giants could have also "devised long-term plans and solutions tailored for the unique context of the Philippines," the report said.
"If the 2016 election was the first time that social media platforms became aware of such manipulations in the Philippines, the election six years later showed that they still lack effective solutions for preventing them," the study said.