TECH

Here's Why Facebook Shuts Down 1 Million Accounts Every Single Day

27/08/2017 4:36 PM AEST | Updated 27/08/2017 4:36 PM AEST

Mark Zuckerberg's social network removes in excess of one million accounts every single day, that breach the user agreement, but is still struggling to stem the tide presented by so-called "threat actors".

This is according to Facebook's security chief, Alex Stamos, who appeared at at conference on Wednesday evening, saying that he believes in the integrity of the spam, fraud and hate speech policies currently in place but they are struggling with the large scale of enforcement.

Stamos said on stage: "The number of accounts turned off every day is (in the) seven figures or eight figures," and later confirmed to CNBC the exact figure.

kasinv via Getty Images

Facebook has long been criticised for its inability to police its own content, after violent deaths and live streams of suicides were broadcast.

Addressing the case of 74-year-old Robert Godwin who was randomly shot dead on a street in Cleveland and it was filmed live, Zuckerberg told a Silicon Valley conference that his platform had "a lot of work to do" and "we will keep doing all we can to prevent tragedies like this."

Then of course there is the ongoing problem of fake news, when Facebook has had to vouch for the 'authenticity' of pages which have been shown to distribute false and misleading news to millions of users on its platform.

Stamos said that the phenomenal reach of the platform, which has approximately 2 billion global users, means implementation is tough.

Not least because in some cases they aren't doing enough, and in others they are criticised for removing posts that are later found to be within the rules.

"When you're dealing with millions and millions of interactions, you can't create these rules and enforce them without (getting some) false positives," he said.

And as the platform operates across more than 100 legal jurisdictions worldwide, Stamos said it is difficult to consistently apply a single version of 'hate speech'.

He explained that the company is planning to expand the team who work on this counter-operation but there is only so much that humans can be expected to do, and some must be delegated to artificial intelligence software.

More On This Topic