Misinformation and disinformation have surged during the current electoral period, drawing attention to how effectively social media platforms respond to flagged content.

Experts from Hashtag Generation and the Institute for Democratic Reforms and Electoral Studies (IRES) have reported a troubling rise in coordinated disinformation campaigns aimed at undermining political figures and manipulating public perception of party manifestos.

This election cycle has also seen an increase in gender-based harassment targeting female candidates, a disturbing pattern that continues from past elections.

Between 15 August and 21 September, the Election Commission (EC) reported a total of 2,010 complaints to social media platforms. 

These complaints include instances of hate speech (713), disinformation (389), using children for campaign purposes (13), misinformation (231), and content contrary to the provisions of the Presidential Elections Act No.15 of 1981 and paid advertisements during the silent period (664). 

Of these, 872 have been removed by social media platforms, while 201 were not removed.

In addition, during the silent period (18-21 September), 1,208 complaints were submitted to social media platforms by the EC, of which 751 were removed. 

Speaking to The Sunday Morning, Hashtag Generation Social Media Specialist Dulanjaya Mahagamage highlighted the effectiveness of Facebook in removing flagged content: “Facebook in particular provided some support and removed some types of content.” 

He added that TikTok had extended similar support, also noting that Hashtag Generation intended to release its report within the next two weeks.

Targeting female candidates

Speaking to The Sunday Morning, IRES Executive Director Manjula Gajanayake said: “There were organised and well-coordinated campaigns of misinformation and disinformation during the Presidential Election period.”

“According to our knowledge and operations, compared to the 2019 Presidential Election, there is minimal hate speech in relation to both the 2024 Presidential and Parliamentary Elections. In 2019, there was hate speech targeting ethnic and religious groups,” he disclosed. 

Mahagamage noted that a significant amount of gender-based violence had been observed during the period leading up to the Parliamentary Elections, especially targeting female candidates. 

He highlighted that this was a continuation of what had been observed during the recent Presidential Election period. “We observed an increase in hate speech and harassment against women during the Presidential Election. This is an extension of that narrative. We have observed a lot of hate speech, misogyny, and sexist comments targeting female politicians and female party supporters,” Mahagamage said.

Gajanayake expressed similar sentiments: “In relation to the Parliamentary Elections, most of the character assassination can be seen in posts made against women. Some of the hate speech and disinformation is shared against female candidates.”

Monitoring and reporting process 

Mahagamage shared that they were yet to notice much disinformation regarding the Parliamentary Elections thus far.

“During the first two weeks after the nominations, the candidates didn’t start their campaigns properly. They started their campaigns only a couple of days ago, so now we have a lot of content. We will get some information when they use their social media accounts,” he added.

He disclosed that Hashtag Generation’s social media monitoring process for the Parliamentary Elections would mirror the one used during the Presidential Election. “We will follow the same procedure, collaborating with the People’s Action for Free and Fair Elections (PAFFREL) and the Election Commission.”

Commenting on the reporting process, he said: “We send cases to PAFFREL and collaboratively send them to the EC. The EC then reports to social media companies.”

However, he pointed out that if there was any extremely harmful content, Hashtag Generation also reported such cases. “If certain content violates community guidelines, then Hashtag Generation has the capacity to escalate it to social media platforms. If this content poses extreme harm, we take quick action. However, we also send this content to the EC so that it can follow up.”

Hashtag Generation reports

In a Hashtag Generation report titled ‘Election Narratives on Social Media’ (1-13 September), its monitoring efforts revealed disinformation surrounding the postal voting process of the recently concluded Presidential Election. 

“Certain Facebook posts claimed that 92% of postal voters voted for Anura Kumara Dissanayake whereas certain Facebook posts claimed that Ranil Wickremesinghe gained 55% of military votes. Both instances amount to disinformation as counting postal votes has not begun yet,” the report read. 

The report also highlighted how social media played a pivotal role in shaping public opinion during the Presidential Election. “We have observed false claims, doctored images, and misleading quotes which can manipulate voters’ views of candidates and policies, sowing confusion and division.”

The report also noted several instances where false claims had been attributed to individuals connected to several political parties. Disinformation concerning one party included fabricated statements, manipulated documents, and misleading claims about the party’s policies and candidates.  

The narratives often attempted to stir religious or ethnic divisions, misrepresent manifestos, and spread false information about economic policies and governance strategies of parties.  

According to the report, there had also been dissemination of false information regarding certain political parties using paid promotions.

In Hashtag Generation’s ‘August Narrative Report,’ there are mentions of politically motivated disinformation campaigns and coordinated and inauthentic social media efforts targeted at discrediting individual associates of a political party. 

According to the report, “harmful content targeting women involved in politics and governance has seen significant growth,” with the type of harmful content varying from “hate speech and harassment to also include misinformation”.