Google said Thursday that it had taken down more than 200 channels weighing in on protests in Hong Kong on its video-streaming site YouTube, the latest technology firm to strike back against Chinese-backed efforts to influence clashes over the future of the region’s fragile democracy.

Google removed the channels after discovering they “behaved in a coordinated manner while uploading videos related to the months-long unrest in Hong Kong,” the company said in a blog post. The online search giant linked the accounts to recent China-backed operations removed by Facebook and Twitter.

Google did not provide information about the content it had discovered beyond saying that those who uploaded the videos sought to disguise their origin — and the company further declined to flag Beijing directly as the culprit.

Google’s move thrusts it into the center of an ongoing debate over tech firms’ responsibility to police their own sites, which is opposed by those who say they should remain largely open forums for free speech. YouTube has typically said much less than other social media platforms about the extent of disinformation it finds and removes from its site.

Twitter earlier this week said it had identified and suspended nearly a thousand suspicious Chinese accounts and banned advertising from state-owned media companies, citing a “significant state-backed information operation” related to protests in Hong Kong. And Facebook said it took down a handful of accounts, as well as several pages and private groups that it said were using deceptive tactics.

“This type of multiplatform activity is not uncommon for these types of influence campaigns,” said Lee Foster, head of the intelligence team investigating information operations for FireEye, a cybersecurity firm based in California. That includes video of “protesters portrayed as being violent amplified as a way to create the impression of widespread criticism of the protests."

Disinformation and other online deceptions have become a focus for the Silicon Valley firms, which are facing a big test leading into next year’s U.S. presidential election. The companies have acknowledged large-scale efforts on their platforms to try to sway voters through phony news articles, videos and forums.

At the same time, the companies have been accused of over-policing some content, drawing criticism for perceived biases. President Trump, for one, has accused Google of suppressing conservative views through its search engine, which Google denies.

The announcements by the tech giants this week also illustrate the reach and pervasiveness of China’s attempts to steer the online conversation about the demonstrations in Hong Kong. State-backed agents have been turning to the largest social media sites in the hopes of swaying protesters in the simmering feud. Social media is cheap or free to use and can effectively reach thousands or millions of people quickly — particularly through mobile phones.

Hundreds of thousands of protesters have filled Hong Kong’s streets following a proposal to allow extraditions to China and other countries, demonstrating in favor of greater autonomy for the region. Though Hong Kong’s chief executive has suspended debate on the bill, it hasn’t been withdrawn completely and she has called the effort “laudable."

The protests forced officials to close Hong Kong’s airport for two days and has sowed chaos as demonstrators clogged main thoroughfares.

Following the first round of takedowns by social media firms, Chinese officials pushed back, saying the accounts were the work of students and others living overseas who should be free to express themselves.

It wasn’t immediately clear what messages the YouTube accounts conveyed or how the company identified them. In its blog post YouTube said noticed the use of private Web networks and “other methods” of subterfuge that are commonplace in influence campaigns.

U.S. officials, including Sen. Mark R. Warner (D-Va.) and Sen. Tom Cotton (R-Ark.), have praised the efforts at culling spurious accounts by the social media giants.