Confident of handling content volume in India, employ 40,000 people for fact-checking: Meta VP Monica Bickert

The company, however, is not looking to replace people from roles that can be performed by technology, she said, adding that better technology did not always lead to fewer people.

“It is not a zero-sum game. Sometimes you need people to label the content to make the technology better. These are two independent things, but they help one another. I would not think of these things as necessarily trading off one another,” she said.

In November 2022, Meta had laid off 11,000 people, or nearly 13% of their total workforce. Earlier this year, it announced plans to lay off another 10,000 employees.

With general elections In India around the corner, the company plans to deploy tools that will identify not just the content but also any trends in the content that starts going viral on Meta’s platforms, such as Facebook and Instagram, Bickert told ET in an exclusive interaction.

“If there’s a certain type of content, something new that is going viral, we can identify that and take a hard look at that to see if it is something that we need to address,” she said.

Discover the stories of your interest

The company employs nearly 40,000 people in India, directly or indirectly, to work on fact-checking misinformation and disinformation. It also has partnerships with 11 fact-checking organisations in the country, Bickert said.The task of removing content, which violates Meta’s community standards, however, becomes more complex with each nuance that is added, she said.

For example, Meta added caste as a “protected characteristic” to flag hate speech. This meant that if a user in India uploaded content that was offensive to others on gender, identity, religion, sexual orientation, and caste among other things, it would be taken down.

“We have to make sure that we are trying to identify the content ourselves. Yes, people can flag to us content that violates our policies. They have always done this. And it is helpful to us. But we do not want to wait for that. We want to find that content before anybody sees it,” Bickert said.

On Wednesday, Meta published its adversarial threat report for the first quarter of 2023 in which it said new malware strains including some posing as ChatGPT browser extensions and productivity tools had been identified.

Over the last two months, the company said it had blocked more than 1,000 malicious links from being shared across its services and shared these links with its peers. The company has also seen an increase in compromise of accounts due to users inadvertently installing malware on their devices, she said.

To help users who may have fallen into such traps, Meta will provide its users with the option of using third-party applications to identify and remove malware and spyware from their devices, Bickert said.

“We are working with security companies to make sure that the search to remove malware is as sophisticated as it can be. But of course, this is going to continue to be an adversarial space,” she said.