Meta's political ad verification, ad library, are efforts to uphold GE15 integrity

Social media giant Meta shared the measures they are taking to foster greater accountability and transparency, as well as combat misinformation for the upcoming 15th general election (GE15) in Malaysia.

This includes increasing transparency for political advertising on its platforms, such as verifying the identity and location of those paying for them, as well as the creation of a publicly-accessible advertisement library.

“Increasing advertisement transparency is an important part of election integrity.

“So, if you want to run political advertisements, you need to get authorised first, then provide information on which organisation you’re from and who is paying for it.

“We also have an advertisement library which allows journalists, researchers and the general public to see who’s behind these advertisements on Facebook and Instagram… see the rough spending and rough demographic these advertisements are targetting.

“This ensures advertisers and us, as a company, are more accountable,” said Roy Tan, who is Meta’s head of politics and government outreach in the Asia Pacific region, at a virtual press briefing today on Meta's efforts in protecting the integrity of GE15.

Tan added that advertisements are archived in their advertisement library for up to seven years, even if they are no longer active.

Tan also addressed why Meta, which owns Facebook, Instagram and WhatsApp, still accepts political advertising even though other tech giants such as Google and Twitter have banned such advertising.

He said they want to foster discourse, especially because social media has become a powerful tool for disseminating information.

Still, at the same time, they want to ensure higher levels of transparency and accountability.

“We based our political advertisement policy on the fact that people should be able to hear from whom they wish to lead them and vote for, and that they should be scrutinised and debated in public,” he said.

Aside from greater advertisement transparency, Meta was also working on detecting and removing any fake accounts created to spread misinformation, whether bots or real people running them.

Tan said the company uses an automated system and machine learning to identify such accounts based on the accounts’ behaviour, and they have more than 40,000 people working on safety and security on their platforms.

In their fight against misinformation, Meta also has policies in place to address harmful content, even if there is no explicit policy saying everything published on their platforms must be true, he added.

They are particularly diligent about removing election-related misinformation that may interfere with people’s ability to vote, he said, such as misinformation about how and where to vote, as well as voter fraud and offers to buy or sell votes.

Meta also works with third-party fact-checkers verified by the Poynter Institute’s International Fact-Checking Network such as, in Malaysia’s case, AFP, Tan said.

“The third-party fact checker will rate whether a story is false or altered and once it is rated as such, we significantly reduce its distribution in our feed so that fewer people see it.

“We also inform people, for example, if something is rated as false, we have an overlay saying this is false but we also give context so people can decide for themselves what to read and what to trust,” he said.

The company also works on disrupting bad actors on its platforms by detecting and flagging harmful election-related activity for manual review and taking action against these suspicious accounts.

Tan said they work in partnership with law enforcement, security agencies and researchers to identify and remove “coordinated inauthentic behaviour” (CIB) networks, which they define as any coordinated network of accounts, pages and groups that rely on fake accounts to mislead users of their services.

Since 2017, 150 CIB networks have been removed, he said.

Limiting misinformation

Aside from that, Tan zeroed in on how they have tried to limit the spread of misinformation on WhatsApp by introducing the forwarded labels and limiting the number of people or groups a message can be forwarded to at a time.

“This keeps the conversations on WhatsApp intimate and personal,” he said.

These measures have resulted in a 70 percent reduction in highly forwarded messages globally, he stated.

Despite the end-to-end encryption on WhatsApp which prevents the company from accessing the content of messages, Meta is also able to ban suspicious users based on “spammy or abnormal behaviour”.

Besides these four pillars, Meta’s head of public policy in Malaysia Noudhy Valdryno highlighted the importance of their community standards, which he described as “the most important guidelines we have in the company”.

He said they have worked on local initiatives to promote an informed and safe election such as capacity-building training for election stakeholders where they are informed on Meta’s community standards and how to use their platforms effectively and accurately.

They have also launched a digital literacy programme in Malaysia called “We Think Digital”, in collaboration with Teach for Malaysia and most recently, Universiti Malaya, he said.

When asked about updates on the alleged links between the police and a troll farm promoting the government and criticising its detractors, Tan said they are constantly on the lookout for “trends and the return of previous bad actors”.

“We take it seriously and will continue to investigate.

“You will only know when the company acts on it and when we announce it because we can’t be too open about it,” he said.

Facebook whistleblower launches non-profit to solve social media harms

US appeals court rejects big tech's right to regulate online speech

Kit Siang: IGP’s reply on ‘troll farm’ claim ‘unsatisfactory’

Cops still probing 'troll farm' allegation – IGP

Expedite probe into police link to troll farms, Guan Eng urges IGP