Mark Zuckerberg said content moderation requires 'nuances' that consider the intent behind a post, but also highlighted Facebook's reliance on AI to do that job

By Katie Canales

Facebook CEO Mark Zuckerberg said the nuances of counter-speech are what makes it difficult to moderate content online, in comments made during a misinformation hearing on Thursday.

By relying on machines to moderate posts, nuance can get lost, a point he made when Democratic congresswoman Doris Matsui of California asked Zuckerberg about anti-Asian hashtags regarding the pandemic that appeared on social media platforms. Specifically, the Facebook CEO said users that make counterarguments against hateful rhetoric sometimes use the same hashtags as people who use them maliciously. 

"One of the nuances that Jack [Dorsey] highlighted that we certainly see as well in enforcing hate speech policies is that we need to be clear about when someone is saying something because they're using it in a hateful way versus when they're denouncing it," Zuckerberg said.

That, in turn, prevents the company from relying on automated moderation to block those words or phrases. 

But Zuckerberg also detailed Facebook's reliance on artificial intelligence in weeding through content online when he was later asked by Democratic Rep. Tom O'Halleran about what the company has done to "increase reviewer capacity."

"More than 95% of the hate speech that we take down is done by an AI and not by a person," Zuckerberg said during the hearing. "And I think it's 98 or 99% of the terrorist content that we take down is identified by an AI and not a person."

Content moderation was a central topic of discussion during Thursday's hearing on misinformation, which was scheduled to discuss the role that Facebook, Google, and Twitter play in how false information spreads online. Misinformation about the COVID-19 pandemic, the 2020 presidential election, and the January 6 US Capitol siege was specifically highlighted throughout the hearing.

The hours-long hearing covered a variety of topics. Tech regulation and issues relating to the industry have become largely politicized, and lawmakers from both sides of the aisle reverted to familiar talking points when grilling the executives.

Some Republicans questioned the CEOs over alleged conservative discrimination online, citing a New York Post article about President Joe Biden's son that was published last year and blocked by Twitter. And Democratic lawmakers grilled the tech leaders over what they said was a lack of action taken to combat misinformation and hate speech online.