No tampering apparent in provincial campaign, Facebook Canada says

The cyberthreat crisis line at Facebook Canada has been quiet during this provincial election campaign, despite a huge reliance on online platforms by parties due to the COVID-19 pandemic.

“We ourselves have not found anything that would suggest that there was this kind of election interference or abuse,” said Kevin Chan, Facebook Canada's head of public policy.

The hotline for parties and politicians was launched following dirty tricks and massive influence campaigns waged during the 2016 U.S. elections by bad actors at home and abroad.

Rather than relying on investigations after the fact, Facebook is trying to battle election interference in real time with a stringent advertising policy that requires multiple levels of identification from anyone trying to run election ads or ads on themes relevant to the election, he said.

The “page transparency” policy adopted by the global personal networking giant allows users to see all the ads run by political parties and interest groups, what they said and the demographics they targeted.

You can also see the ads that were pulled out of circulation for breaking the rules.

“You can see the advertisements in the ad library that were pulled because it didn't have a proper disclaimer, meaning they didn't go through the authorization process,” Chan said at a briefing for reporters. “That isn't evidence of abuse, it might just be an innocent mistake.”

Facebook pulled up its socks on policing misinformation and co-ordinated campaigns to manipulate voters after being ensnared in a scheme by Cambridge Analytica to harvest users' personal data for use in political campaigns. The campaigns for American politicians Ted Cruz and Donald Trump employed Cambridge Analytica.

The platform now strictly limits how much information third parties and app builders can collect from users, “even with their consent,” Chan said.

Facebook employs sophisticated tools to identify fake accounts and bad actors engaging in “co-ordinated inauthentic behaviour.”

So far this year, they have removed 3.2 billion fake accounts, most of which are generated through automation and are detectable because they don't show the hallmarks of normal human interaction with the platform.

“If you are on Facebook, people are expected to be who they are and who they say they are,” he said.

“We are using technologies that didn't exist a few years ago to proactively find fake accounts and remove them en masse, ideally before they are even used.”

Fake accounts comprise about five per cent of active users on the platform, according to the company's quarterly enforcement report.