The U.S. military created and operated fake Arab-language Twitter accounts that pushed U.S. interests in the Middle East and Twitter “whitelisted” those accounts, according to a new report from the Intercept.
The report, published yesterday, expands on reporting done by online security researchers affiliated with the Stanford Internet Observatory. In August, the researchers reported on what they suspected to be thousands of accounts created and run by the U.S. government as a part of a state-backed information initiative. Many of those accounts used photos for their profile picture that were created using artificial intelligence, or “deep fakes.”
Yesterday, the Intercept reported that emails it obtained showed that at least one of the accounts was directly affiliated with the Pentagon.
One account, @mktashif, originally disclosed its affiliation with U.S. Central Command (CENTCOM) in its Twitter bios but at some point, the clarification was deleted, according to the Intercept. Around the same time, the accounts’ photo was also changed to one that the Stanford researchers identified as a deep fake.
The account’s new bio claimed, in Arabic, that it was an unbiased source of news and information and “dedicated to serving Iraqis and Arabs.” It regularly tweeted messages that aligned with U.S. interests, like denouncing Iran and Houthi rebels in Yemen.
According to the report, Twitter was aware of these accounts and their use by CENTCOM. In July 2017, a CENTCOM official emailed a Twitter public policy representative asking for one account to be approved for verification and to “whitelist” a list of Arab-language accounts that the Pentagon used “to amplify certain messages.”
CENTCOM sent a list of 52 accounts, ranging from @yemencurrent, which tracked U.S. drone strikes in Yemen and claimed they were “accurate” and killed terrorists, to accounts that spread anti-Iran and Iraq messages, as well as ones promoting U.S.-backed militia groups in Syria. According to a Twitter engineer who spoke to the Intercept, “whitelisting” those accounts essentially gave the accounts the privilege of being verified on the app without the visible blue check.
Twitter has said for years that the company would take down state-backed accounts but never made an explicit exception for the U.S. The emails published by the Intercept show that not only was Twitter fully aware of the state-backed accounts, it granted them privileges at the explicit behest of the military.
The report Tuesday is just the latest in a series of reporting on Twitter from what has been dubbed the “Twitter Files.” New Twitter CEO Elon Musk has allowed a select few journalists access to Twitter’s emails and records for short amounts of time and has encouraged them to “surface anything bad Twitter has done in the past.”
