March 20, 2012

Twitter bots — zombie accounts that auto-follow and send junk tweets hawking questionable wares and services — can be an annoyance to anyone who has even a modest number of followers. But increasingly, Twitter bots are being used as a tool to suppress political dissent, as evidenced by an ongoing flood of meaningless tweets directed at hashtags popular for tracking Tibetan protesters who are taking a stand against Chinese rule.

It’s not clear how long ago the bogus tweet campaigns began, but Tibetan sympathizers say they recently noticed that several Twitter hashtags related to the conflict — including #tibet and #freetibet — are now so constantly inundated with junk tweets from apparently automated Twitter accounts that the hashtags have ceased to become a useful way to track the conflict.

The discovery comes amid growing international concern over the practice of self-immolation as a means of protest in Tibet. According to the Associated Press, about 30 Tibetans have set themselves on fire since last year to protest suppression of their Buddhist culture and to call for the return of the Dalai Lama — their spiritual leader who fled during a failed 1959 uprising against Chinese rule.

I first heard about this trend from reader Erika Rand, who is co-producing a feature-length documentary about Tibet called State of Control. Rand said she noticed the tweet flood and Googled the phenomenon, only to find a story I wrote about a similar technique deployed in Russia to dilute Twitter hashtags being used by citizens protesting last year’s disputed parliamentary elections there.

“We first discovered these tweets looking at Twitter via the web, then looked at TweetDeck to see how quickly they were coming,” Rand said in an email to KrebsOnSecurity.com late last week. “They no longer appear when searching for Tibet on Twitter via the web, but are still flooding in fast via TweetDeck. This looks like an attempt to suppress news about recent activism surrounding Tibet. We’re not sure how long it’s been going on for. We noticed it last night, and it’s still happening now.”

Denis Sinegubko, the Russian security blogger I corresponded with during the twitter bot campaign against anti-Kremlin activists last year, said the Twitter accounts being used to flood Tibetan hashtags have all the hallmarks of Twitter bots.

“A brief analyses of the ‘accounts’ that use these hashtags suggests that they are spam bots: too many messages with the same hashtags in a very short time – unlikely to be humans,” Sinegubko said.

Twitter was very responsive to the botted accounts being used to drown out hashtags following the disputed Russian elections, but these anti-Tibetan twitter bots appear to have flown under the radar so far.

When I checked the situation Monday evening, the bunk tweets aimed at popular Tibetan hashtags were still going strong. It’s not immediately clear how many apparently botted accounts are being used to blast these tweets; most of them have zero — if any — followers, and are following very few other accounts. Twitter has been notified about a couple of dozen accounts that appear to be the source of most of these junk messages.

Update, 1:44 p.m. ET: Several security firms are reporting that a backdoor Trojan targeting Mac users via a Java vulnerability has been observed in email attacks against non-governmental organizations related to Tibet. More information from SecureMac and AlienVault.


10 thoughts on “Twitter Bots Target Tibetan Protests

  1. Jan Doggen

    So there is now a market for anti-spam solutions for Twitter. Anyone know of any apps in that field?

  2. Blackbeard

    Please don’t give the Republicans any ideas!

  3. -B

    Please don’t let the Democrats politicize this. It’s cross-party in nature. Vote Libertarian!

  4. Silemess

    Really? We’re going to go off topic about politics ourselves?

    Is Twitter serious about closing the new round of bots? Russia wasn’t as serious a market for their attempts to enter. China’s restrictions means that Twitter wants to play nice so that it can be invited to come over and play. So will they move against the bots as effectively as they did for Russia?

    I suppose the solution is to build a filtering system that grabs up hashtags from white listed accounts with a separate box for the unlisted tags. Building up both a white and black list would allow them to give new voices a chance, but is a more intensive program than simply approving known good sources.

  5. @Buxoro

    This has been going on for a while. I tried to contact Twitter support many time with no response. From what I can see, it’s all started one year ago, when the Tibetan situation started getting worse, the intensity of the attack increases.

    Most of these are for one Tibetan twitter user living in China, her got mentioned up to 20 thousands a day, and is closely correlated with the #tibet and #freetibet tags.

    I really what to find a way for her to automatically identify the twitter accounts and block those more efficient than what she’s doing now, i.e., manually and one by one.

  6. uzzi

    “Flooding” is so clumsy – maybe the chinese should adopt to western standards: Search “Four Minute Men” on Wikipedia and improve to a budget of $80 billion/year… having propagandistic ghostwriters brainwashing people everywhere looks as stupid as flooding but feels more civilized. ;-P

    1. @Buxoro

      Thanks for that! I’ve tried to run it on my Mac, looks promising but kinda slow.
      Do you know if there is a way to report spam (and block) the spammer automatically with the script?

Comments are closed.