Representatives from Facebook, Google and Twitter testified Tuesday in front of a United States Senate committee on activity of Russian entities using the platforms to spread disinformation before, during and after the 2016 U.S. presidential election.

The Senate Committee on the Judiciary, Subcommittee on Crime and Terrorism—chaired by Senator Lindsey Graham, R-SC—grilled the witnesses sent by the tech firms on a variety of issues ranging from Russian interference, the undue influence held by the tech firms and the presence of extremist content being spread across social media platforms.

During the session, titled "Extremist Content and Russian Disinformation Online: Working with Tech to Find Solutions," the tech firms revealed details about the extent of Russian activity on the respective platforms, which was more significant than originally reported.

According to the written testimony submitted by Facebook General Counsel Colin Stretch, the social network found a Russian government-funded troll farm known as Internet Research Agency (IRA) spent approximately $100,000 on more than 3,000 Facebook and Instagram ads between June 2015 and August 2017.

Those ads, which the company claimed were used to promote the roughly 120 Facebook Pages run by Russian groups that produced posted more than 80,000 pieces of content between January 2015 and August 2017.  Facebook estimated that 11.4 million people in the U.S. saw at least one of the IRA-placed ads between 2015 and 2017, with 44 percent of the impressions coming before the 2016 presidential election.

That information spread farther as users on the platform interacted with it. According to the company, content from IRA-linked accounts were served to roughly 29 million people through their newsfeeds, from which users shared, liked, and followed accounts originating from the Russian troll farm. Facebook said its “best estimate” is approximately 126 million people saw content from an IRA-created account at some point between 2015 and 2017.

Dianne Feinstein, D-CA, revealed in her opening statement before the hearing that Facebook informed members of the Senate committee that it identified 470 accounts tied to IRA on its platform.

”We take what happened on Facebook very seriously. The foreign interference we saw is reprehensible," Stretch said before the committee.

Sean Edgett, the Acting General Counsel for Twitter, testified that IRA was similarly active on its platform. According to Edgett’s testimony, Twitter has so far identified 2,752 accounts tied to the Russian troll farm during its ongoing investigation—a considerable increase from the 201 originally reported to the committee.

Those IRA-run accounts, many of which were automated in their behavior, posted more than  131,000 tweets in the lead up and direct aftermath of the 2016 presidential election, though Twitter reported just nine percent were directly related to the election itself.

Google revealed arguably the smallest presence of Russian activity across its platforms, which the company’s Senior Counsel for Law Enforcement and Information Security Richard Salgado suggested was because information does not spread on Google in the same viral manner that it does on other social networks.

Google revealed in its written testimony that it found two accounts on its advertising platforms that it believed to be backed by government-sponsored entities. Those accounts spent roughly $4,700 on advertisements in connection with 2016 presidential election.

The search giant also said it found 18 YouTube channels that it believed to be associated with Russian government-sponsored organizations. Those channels produced a total of 1,100 videos—about 43 total hours of content—of which just three percent amassed more than 5,000 views.

All three companies attempted to downplay the reach of the content from Russian-backed agencies and operations, noting the content produced by those groups accounted for just a fraction of a percent of the total content that appeared on the platforms.

The argument did not move many of the senators, who pressed the companies to take additional action to identify foreign influencers attempting to influence U.S. elections and sow political discord among citizens.

Perhaps the tensest moment of the hearing came when Senator John Kennedy, R-LA, placed Facebook in the crosshairs and questioned the service’s ability to keep track of every advertiser using its platform.

"You've got five million advertisers, and you're going to tell me that you're able to trace the origin of all of those advertisements?" the senator asked, pressing Facebook’s representative on whether the site knew if North Korea, China or others may have placed ads during the 2016 presidential election like Russia did.

Facebook General Counsel Stretch admitted there were steps it could take to track the origins of some advertisers but ads placed by funneling money through shell companies would make it difficult for the site to ensure the true identity of the entity paying for an advertisement. Stretch also conceded the site did not know if other nation state actors used the ad platform to spread information during the 2016 election.

Senator Kennedy, who said told the tech giants “your power sometimes scares me,” also took a portion of his five minutes of questions to get the companies on the record on a number of points of contention about how they operate.

Kennedy asked Facebook’s general counsel if the company’s CEO could ask an employee of the site to surface “everything we can find out about Senator [Lindsey] Graham, he could do that couldn’t he?”

Stretch paused before insisting that Facebook has designed its systems to “prevent exactly that” kind of abuse despite the service maintaining massive amounts of information about individual users and allowing advertisers and companies to target individuals based on specific details.

Kennedy also asked Google if it was a media company or not, noting that its platform is a massive distributor of information for many people and acts similarly to a newspaper. Google’s Salgado insisted Google is a tech company and said “we are not a news platform.” Kennedy responded by saying, “That’s what I thought you’d say,” before ending his questioning.

The tech firms also faced a stern line of questioning from Al Franken, D-MN, who lamented the fact that the companies have access to massive wells of data and can offer hyper-targeted advertising but couldn’t figure out when activity was coming from foreign actors.

“How does Facebook, which prides itself on being able to process billions of data points and instantly transform them into personal connections for its users, somehow not make the connection that electoral ads, paid for in rubles, were coming from Russia?” the Senator asked.

Stretch attempted to admit fault by saying the company should have applied a “broader lens” to how it scrutinizes content, but Senator Franken interrupted to say that tech companies have “all knowledge that man has ever developed,” before asking, “You can’t put together rubles with political ads and go, ‘Hmm, those two data points spell out something bad?’”

Much of the hearing focused primarily on advertising and paid content spread by Russian government-backed organizations, an issue for which a legislative solution has been proposed.

The Honest Ads Act, submitted introduced this month by a bipartisan group of Senators, would create stricter rules and increase transparency for political advertisements placed online. The proposal, or at least its primary intention, was supported by the tech companies when asked during the hearing.

Little of the hearing focused on the broader issue of foreign interference and propaganda spread across social media platforms for free and without much effort from adversaries—an issue that stems from fundamental problems within the social networks, from algorithmic models that propel popular stories regardless of its basis in fact. Those issues may not have a simple, legislative solution.