More fact-checking, dedicated teams: Facebook details strategy to curb misinformation during polls
The social media giant said it is blocking or removing about one million accounts a day.
Facebook on Monday detailed its efforts to combat fake news and the spread of misinformation ahead of the Lok Sabha elections in India.
In a blog post, Ajit Mohan, the company’s Managing Director and Vice President, India, listed out the steps that the social media giant has been taking in recent months to develop more robust mechanisms to prevent the misuse of its platform. These include measures to increase transparency in political advertisements, tie-ups with third-party fact checkers to detect fake news and coordination with the Election Commission of India.
“We are committed to working hard to prevent abuse on our services, especially during elections,” Mohan said. He said that dozens of teams had been working around the globe for more than 18 months to develop strategies.
“The findings allowed us to concentrate our work on key areas, including blocking and removing fake accounts; fighting the spread of misinformation; stopping abuse by domestic actors; spotting attempts at foreign meddling; and taking action against inauthentic coordinated campaigns”, he said.
A key measure, the post said, was the launch of political advertisement transparency tools last year. All political advertisements are accompanied by “paid for by” and “published by” disclaimers. There’s also an Ad Library which has a collection of all political advertisements on the platform along with data of who has seen them.
Facebook has also introduced new regional operations centres comprising engineers, specialists and data scientists, “focused on election integrity”, in Singapore and Dublin. The teams will work with local experts in Delhi and staff at Facebook’s California headquarters. A report in The Hindustan Times said that the operations centres comprise 40 teams of 30,000 people who keep track of material published on Facebook to see if anything harmful, misleading or otherwise in breach of their code of ethics is being shared.
Using artificial intelligence and machine learning to fight interference, Facebook said it has been able to block or remove about one million accounts a day, ostensibly those that the company suspects are publishing content that is misleading or goes against their guidelines.
Fact checking and more
Mohan’s blog said that the company has also tied up with seven accredited organisations in India for fact-checking services. Content in English, Hindi, Bengali, Marathi, Telugu, Tamil, Malayalam and Gujarati will be covered under these partnerships.
On how this system would work, he wrote, “When fact-checkers find a story to be false, they’ll write an article that reports on the actual facts at hand. We include these articles as Related Articles immediately below a story in News Feed. We also show the debunked articles lower in people’s feed.”
Notifications will be sent to people and page administrators if they share a story marked as false.
Along with other social media firms, Facebook has agreed to a “voluntary code of ethics” for the General Elections with the Election Commission. According to the code, which came into effect on March 20, the poll body can inform the relevant social networking site of potential violations of election laws and guidelines on their platform. They will then need to act on these violations in a time-bound manner.
Political advertisements need a pre-certification from the Election Commission or the Media Certification and Monitoring Committee. The 48-hour blackout period on political campaigning before voting day will also apply to social media platforms.
“We’ve also created a training process to help policymakers, candidates and their staff improve their cybersecurity and awareness for how their accounts could be hacked or abused,” Mohan wrote.
Facebook has been under pressure to build more robust mechanisms to track misinformation on the platform after allegations of inadvertently enabling Russian meddling in the 2016 US elections.
Last week, the social media giant announced that it it had removed about 700 pages in India that it claimed were displaying “coordinated inauthentic behaviour”, which the platform describes as a concerted effort by a large group to mislead others. This included several pro-Bharatiya Janata Party and pro-Congress pages.
Also read:
- Indians are being scammed on Facebook with Modi’s photographsand the lure of cheap smartphones
- Pro-BJP pages account for 70% of ad spending made public by Facebook, analysis shows
- As pro-BJP sites lead on Facebook ad spending, questions of transparency emerge
- Pro-BJP Facebook pages offer freebies for votes, raising questions of poll code violation
- Are WhatsApp, Facebook, and Twitter ready for the Indian election?