Twitter mounts robust fightback against Chinese propaganda outlets

<p>
<strong>Pushkar Sinha</strong></p>
<p>
Twitter, the social media giant, has stunned the Chinese propaganda machine. Recently, it has gone the extra mile to pick up Chinese state-sponsored tweets, and alerted users about the mischief. Using sophisticated algorithms, Twitter has suitably stamped these tweets as products of “CHINESE STATE AFFILIATED MEDIA”.</p>
<p>
The appellation has gone a long way in alerting Twitter users that they are encountering propaganda material marshalled by the Chinese authoritarian state. As a result, a sharp decline in viewership of these tweets has been registered.</p>
<p>
Twitter could embed suitable filters, only after fully understanding the Chinese modus operandi for expanding Beijing’s propaganda footprint.  </p>
<p>
<iframe allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen="" frameborder="0" height="315" src="https://www.youtube.com/embed/z7CPqROtanA" width="560"></iframe></p>
<p>
According to a research report published by ProPublica, a New York based non-profit organisation that publishes investigative journalism in public interest, the Chinese web of deception to hijack valid Twitter accounts begins fairly innocuously. The first step is to funnel content in Mandarin language in a person’s Twitter Account. The subscriber, who is located away from mainland China, at this point, assumes that his or her account has been hacked. Consequently, the person opens a replacement account.</p>
<p>
Phase two of the operation begins now, after the valid account has been hijacked.</p>
<p>
With the follower base already built by the previous user, the valid account then begins tweeting propaganda steered by the Communist Party of China (CPC).</p>
<p>
The subjects used in the hijacked accounts can vary. For instance, the account can be used to promote the China made vaccine, while deriding and questioning the safety of West made jabs to counter coronavirus. Theories of the Covid-19 pandemic starting from somewhere other than China are floated. The Hong Kong Police handling of the protests is praised. By the end of the month, the account gets a Chinese name and bio and the transformation is complete.</p>
<p>
The ProPublica report further states that a thorough examination of interconnected accounts suggests that OneSight, an internet marketing company is covertly pulling the strings of the operation.</p>
<p>
Records indicate OneSight holds a contract to boost the Twitter following of <em>China News Service </em>- one of the largest state-owned news agency. Some of the bot accounts owned by this internet marketing company are eerily similar to legitimate <em>Western News Agency</em>’s Twitter accounts. Masquerading as a news agency helps provide an element of authenticity to the propaganda spread by these accounts.</p>
<p>
By 2019, Twitter began its fightback, taking concrete steps to combat the misuse of its platform. In this year, it enacted a policy to stop state-linked accounts from promoting and advertising tweets on their plank. Following through with its decision it suspended around 5,000 accounts suspected of being controlled by the Chinese state, and banned 2,00,000 related accounts that are not very active. It also attached a ‘Chinese state affiliated media’ to accounts run by official mouthpieces of the party and the journalists working for them.</p>
<p>
Tweets by newspapers such as <em>China Daily </em>and CPC mouthpiece <em>People’s Daily</em>, global broadcaster <em>CGTN </em>and <em>Xinhua</em>, an official news agency, now appear with the same. Their tweets have also been restricted from being displayed as the ‘top’ result among searches.</p>
<p>
In doing so, Twitter was attuning with Facebook, which had a similar policy in place from June 2020. According to a report by Stanford's Cyber Observatory Centre, labelling tweets has had a significant impact on the reach of the tweets-likes and retweets have declined by 20% and 30% respectively.</p>
<p>
Advanced tools like a Machine Learning Algorithm have been deployed by Twitter to identify illegitimate accounts. Engagement with these fake accounts has also hit an unprecedented low. An analysis of 35,00,000 tweets by Amal Sinha, an independent researcher suggests that 98.22% of the posts do not receive any likes and retweets. Recently these accounts began tweeting pornography in an attempt to increase engagement.</p>
<p>
A report published by Jordan Schneider, an independent researcher and data scientist based in New York makes it evident that the people controlling these accounts are given weekends off. Data suggests an average twitter user increases their engagement during the weekends while the bot accounts cease tweeting.</p>
<p>
There is also evidence that suggests that tweet volume by these accounts increases during Chinese standard time, even reducing during their lunch hours. This feature signals that the propaganda tweets are likely to have a Chinese origin.</p>
<p>
The bot accounts also lack political nuance. For example, accounts claiming to live in Hong Kong usually write in Simplified Chinese Characters which is widely used in China but not Hong Kong.</p>
<p>
With its propaganda failing to have an impact on social media, the CPC has now set its sight on popular Chinese Youtubers, using its carrot and stick policy to influence the content. There was an attempt to rope in Youtubers such as Winston Frederick Sterzel, also known by his YouTube pseudonym SerpentZA– a South African vlogger and video producer who lived in Shenzhen in the Guangdong province of China for 14 years—and Matthew Tye in the project. Their refusal to accept the CPC’s request was met with relentless harassment, prompting them to leave the country.  Finally, Sterzel moved to Los Angeles in 2019.</p>