Facebook chief operating officer Sheryl Sandberg and Twitter CEO Jack Dorsey appeared before the Senate Intelligence Committee Wednesday morning to answer a range of questions about the social media companies’ policies, their handling of disinformation and how they were addressing bots and other “bad actors” on the platforms. Representatives for Google did not attend after declining to send a high-ranking executive.
The questions from both Republicans and Democrats ranged from tedious to contentious. Executives discussed and defended their companies’ data privacy rules, their responses to disinformation campaigns and their approaches toward removing hate speech. Several lawmakers, including Sen. Mark Warner (D-Virginia), the Senate committee’s vice chairman, said regulation was coming.
Here are some of the major takeaways from Sandberg’s and Dorsey’s comments during Wednesday’s panels.
Facebook and Twitter still don’t know how to deal with bots
Both Sandberg and Dorsey conceded Wednesday that bots constituted a major problem that their platforms were still struggling to address. Sandberg said that while Facebook’s rules prohibiting fake accounts were clear-cut, the company had more trouble identifying them.
Dorsey, whose answers throughout the day were less rehearsed than Sandberg’s, said he believed it was “useful” to label certain accounts as automated, but said the benefits of that were limited to how accurately Twitter’s systems are able to identify those accounts. He also suggested that Twitter’s imperfect ability to identify bots could lead some users to believe that accounts not labeled as bots were human-run.
One obstacle to identifying and removing bots is that sophisticated actors can script a bot’s actions to make it appear human-like, Dorsey said. Dorsey said Twitter challenges between 8 million and 10 million suspicious accounts every week to prove they’re human and prevents another 500,000 every day from logging in to the platform.
Among the senators concerned with bots and foreign interference was Susan Collins (R-Maine), who was targeted by the Russia-linked Internet Research Agency. She said tech companies like Facebook and Twitter need to do a better job of informing users who are targeted by bots or disinformation campaigns, a point that both Sandberg and Dorsey agreed with.
Combating disinformation falls to outside actors
In the morning hearing, Sandberg highlighted Facebook’s partnerships with third-party fact-checking organizations as the best way to handle disinformation. Dorsey said that Twitter needed to do a better job of amplifying verified journalists on the platform who debunk false information.
During testimony highlighting the benefits of collaborating with researchers, Sandberg pointed out that a private security firm helped identify Iranian actors in the most recent revelation of foreign interference.
The testimony showed that both platforms are still relying heavily on third-party groups and individuals to handle the vast disinformation issues rather than taking an in-house approach. Dorsey said relying on its own users to report violations was a “matter of scale” and conceded that the platform needed to do more to combat harassment. But the details of what “more” will look like in practice remained unclear.
Sen. Kamala Harris (D-California) and Sandberg had a tense exchange during which Harris asked Sandberg about policies against hate speech on the platform. She referenced a 2017 ProPublica report that found that groups like black children could be disparaged on Facebook because they were not counted as a “protected class.”
Sandberg said that Facebook had since changed what she said was a “badly written policy.” In an email to Adweek, a Facebook spokesperson said the original definition was “too narrow” and that the platform’s community guidelines now take more ambiguous attacks against different groups of people into consideration when enforcing its rules.