The Facebook clean-up attempt on political messaging

How is Meta planning to moderate its content to deal with misinformation and fake accounts? What are the new rules for advertisers on the site?

The story so far: In a blogpost on March 27, Facebook’s parent company Meta stated that it can anticipate threats and help prevent interference in elections better than before. It added that the platform’s focus on strengthening artificial intelligence and machine learning systems among other tools made it more effective at finding and removing abuse and fake accounts. Further, to ensure a level playing field, Meta extended the use of all advertising tools, previously available only to large entities, to entities of all sizes.

The Ad auction: Meta said ad pricing on its platforms vary as per the entity’s targeting and bid strategy. It informed the auction does not make pricing decisions based on political viewpoints. “The assertion that any political party in India got discounted rates on ads because of their political affiliation is factually inaccurate,” the blog stated. In January, Facebook decided to do away with some of its ‘detailed targeting’ options that let the advertisers refine their target audience. It informed that broad targeting along with customised and lookalike audience targeting options would continue to exist.  

Disclaimers: Political ads on Facebook and Instagram are required to be mandatorily authorised and must include disclaimers as well. This enables users to note the name of the person or organisation running the ads. The advertisers are required to verify their credentials to increase their accountability. “For example, if we discover that the phone, email or website are no longer active or valid, we will inform the advertiser to update them. If they do not, they will no longer be able to use that disclaimer to run ads about elections or politics,” the blog read.  

Removing and demoting content: Meta announced that all political advertisements violative of their policies and standards, including those flagged by the Election Commission of India and scrutinised by them, will not be allowed to stay on their platforms. Separately, prior to the recently concluded State elections in five States of India, it had stated that if content, even though not violative of its standards but could potentially cause offline harm becomes widespread, it will be demoted so fewer people can see it. 

Coordinated Inauthentic Behaviour: This refers to coordinated actions at a large scale aimed at influencing public opinions to achieve a strategic social or political goal. Fake accounts are usually central to such operations. Meta stated it accords focus on behaviour and not content when determining deceptive campaigns. “...in many cases the content shared by influence operations isn’t verifiably false and may in fact be copied from authentic communities these deceptive campaigns are trying to mimic or reach,” the blog stated. The blog informed that it has been working on technologies to detect and block fake accounts, which are the source of a lot of inauthentic activity.  

Virality: Meta said it does not tweak its algorithms to suit a particular user. The feed is essentially shaped by choices and actions of the individual user. “It is made up primarily of content from the friends and family they choose to connect to on the platform, the Pages they choose to follow, and the Groups they choose to join. Ranking is then the process of using algorithms to order that content,” the blog stated. 

Language: Meta’s blogpost stated that it has content reviewers in 20 Indian languages. There has been $13 billion worth of investment which helped the company triple the size of its global team working on safety and security to over 40,000 including 15,000+ dedicated reviewers across 70 languages.