People holding mobile phones are silhouetted against a backdrop projected with the Twitter logo in this illustration picture taken in Warsaw. PHOTO: REUTERS
The released data is part of an internal investigation by the social media giant into allegations of foreign meddling on their website, according to a statement released by the company in a blog post.
In September, Twitter chief Jack Dorsey had appeared before a US Senate Intelligence Committee in Washington to brief lawmakers regarding efforts the company was making to combat coordinated misinformation campaigns.
The Senate hearing was probing allegations of foreign interference in the 2016 US presidential elections, which had surfaced in the media just before Donald Trump was elected the 45th President of the United States.
During the hearing, Dorsey had assured public representatives and the media that Twitter would publicly provide updates of the internal inquiry on a regular basis.
Earlier in January, the company released a statement saying it had identified a Russian government-linked organisation, called Internet Research Agency (IRA), which operated 3,814 accounts involved in a 'propaganda effort' to target the US elections. However, Twitter had not released any other information related to the accounts at the time.
Recently released figures provide in-depth details about these accounts, as well as 770 other accounts, which the company said had originated in Iran. The new files include more than 10 million tweets and more than 2 million images, videos and other media.
The Express Tribune has performed an analysis on parts of the dataset shared by the social media website.
Pakistan is mentioned in 5,652 tweets in some documents related to suspect accounts originating from Iran.
Platforms used to post tweets
[jtrt_tables id="1828676"]
Data archives of accounts analysed revealed that the web client of the social media giant was used most frequently for posting content on the website. Third-party applications and media managing tools were also used to spread disinformation, including Facebook and Buffer.
According to Twitter, the locations with which the tweets were identified had been the 'self-reported locations' that the users had posted.
Interestingly, most of the content flagged as suspicious was apparently coming from Brazil. France, Turkey, Iran and the US were also featured on the list.
Websites shared in Twitter posts
[jtrt_tables id="1828699"]
The archive showed websites that suspected users had tweeted about most frequently. According to figures, AWD News was on top of this list, with whatsupic and libertyfrontpress following close behind.
A number of these websites, including AWD News, have been already identified as platforms used to generate fake news.
URLs shared on Twitter
[jtrt_tables id="1828705"]
Twitter released an accompanying list of web addresses that were shared widely by users with intent to spread misinformation.
These links are mostly from AWD News and libertyfrontpress, and can help the public understand why certain stories are shared more than others by suspect groups.
The Express Tribune went through some of these links in order to separate fake stories from real stories.
http://www.awdnews.com/political/10404-cia-predict-third-terrorist-attack-after-sidney-and-pakiistan-in-usa-in-3-days.html
Headline: CIA predict third terrorist attack after Sidney and Pakistan in USA in 3 days
The link shared above is no longer available. British newspaper The Guardian has already identified AWD as a fake news site. Wayback Machine screenshot shows that the article is full of spelling mistakes and the story has not been credibly sourced. The URL reproduced above was published on December 16, 2014, and tweeted 3,619 times.
http://www.7sabah.com.tr/haber/6876/pakistan-genelkurmay-baskani-israili-12-dakikada-yok-ederiz/
Google translation of headline: Pakistan Chief of Staff: We are not in Israel in 12 Minutes!
The article linked above shows pictures of former Chief of Army Staff General (retd) Raheel Sharif, and most sites quoting this also claim Joint Chiefs of Staff General Zubair Mahmood Hayat spoke to AWD News. This URL was published on October 31, 2016, and tweeted 72 times.
http://whatsupic.com/news-politics-world/1476905223.html
Headline: We can destroy Israel in ‘less than 12 minutes’: Pakistani commander
This story was tweeted out 67 times and originally published on October 19, 2016.
http://www.awdnews.com/index/saudi-arabia-bought-depleted-uranium-weapons-pakistan-delivered-syrian-rebel-groups/
Headline: Saudi Arabia has bought depleted uranium weapons from Pakistan and delivered it to Syrian rebel groups”
The story linked above was tweeted 47 times and originally published on January 14, 2018.
http://www.libertyfrontpress.com/70-american-security-services-express-fear-of-pakistans-nuclear-program/
The article linked above was tweeted 7 times, and the link to it has stopped working at the time of writing this article.
The Washington Post quoted cyber-security firm FireEye as stating that libertyfrontpress is an influence operation masquerading as liberal US activists — and apparently aimed at promoting Iranian political interests.
[http://sachtimes.com/en/pakistan/19065-saudis-direct-role-in-nawaz-sharif-dismissal-leaked-letter-revealed/]
Headline: Saudis direct role in Nawaz Sharif dismissal, leaked letter revealed
The link above has been removed and was tweeted twice, according to data reviewed by The Express Tribune.
Hashtags used to spread select news
[jtrt_tables id="1828709"]
Archived data revealed that certain hashtags were used to spread select news items on Twitter, mostly through a specific group of individuals.
Certain keywords, like Pakistan, US, Trump, Syria, Afghanistan, and bombs were added to stories in order to maximise their reach.
Records kept by the social media website show that Twitter users identified as suspect had created their profiles on the platform most frequently in 2014 and 2011.
All URLs mentioning Pakistan
[jtrt_tables id="1828725"]
COMMENTS
Comments are moderated and generally will be posted if they are on-topic and not abusive.
For more information, please see our Comments FAQ