Twitter last week said it suspended 936 accounts from the Chinese mainland for allegedly disseminating disinformation on the situation in Hong Kong. Facebook and YouTube quickly followed suit, closing down accounts which they claim had violated community standards.

"These accounts were deliberately and specifically attempting to sow political discord in Hong Kong, including undermining the legitimacy and political positions of the protest movement on the ground," wrote Twitter in a statement.

Twitter's move came against a public outcry over social media platforms' lack of transparency about the limits of free speech.

Some voices are fearing social media would silence them. Algorithms that decide what users see and with whom they interact are creating filter bubbles leading to less common acknowledgement on what "facts" are.

One account listed by Twitter as an example of spreading "disinformation" had published pictures of violent protesters smashing their way into and trashing the Legislative Council building on July 1. The pictures were accompanied by the message: "We don't want radical people like these in Hong Kong. Get out of here."

The vandalism and violence in the Legislative Council were widely reported by international media. It is not clear what standards Twitter follows when deciding what constitutes "disinformation."

"We have a problem in how we deal with disinformation in general," said Shaun Riordan, director of the Department for Diplomacy and Cyberspace at the European Institute of International Studies, at a recent symposium held by the Charhar Institute in Beijing.

Having worked with the EU task force, which focused on combating disinformation from Russia, he said one of the central problems is that "they are bad at distinguishing disinformation and different political perspectives." For example, on occasions, materials that simply offer a different perspective may be regarded as fake news.

Riordan suggested that social media giants either take down accounts that encourage demonstrating in Hong Kong along in step with banning the 936 anti-protest accounts, or refrain from hushing everyone. "They are taking a role which is not theirs and they are taking sides in the discussion."

Other than alleging that the accounts spread "disinformation," Twitter claimed that the accounts were suspended for a number of violations of the company's platform manipulation policies, including spam, coordinated activity, fake accounts, attributed activity and ban evasion.

But according to a paper published in 2018 by researchers from Oxford University, no evidence was found of pro-Chinese-state automation on Twitter.

On the contrary, the study found that information about China and Chinese politics on Twitter has been dominated by anti-China voices, with a significant number of these users being automated or fake accounts. Among the top 100 most active users posting China-related information on Twitter, none posted pro-government content and half of them posted anti-government information.

"This analysis suggests that the Chinese state is not utilizing automation to influence discourse on Twitter… it is anti-state groups with few resources who are using automation to manipulate information about China and Chinese politics," the study read.

A Hong Kong resident, whose Facebook account was taken down purportedly because his profile photo showed his support for the Hong Kong police, told CGTN over the phone, "Right now, most of the residents [in Hong Kong] believe the police did a lot of bad things. That is not fair. And that is why many of us are trying to voice our opinions through social media."

The individual, who requested anonymity, added: "We did not get money from the Chinese government... we did what we believe we should do as Hong Kong residents."

Facebook, Twitter and YouTube, by taking sides on political issues, risk losing their credibility as platforms open to diverse voices and turning them into echo chambers. In the marketplace of ideas, different sides are supposed to fight it out and let truth emerge in the process.

Unfortunately, by blocking accounts that offer a different perspective, these social media giants are preventing that from happening.