Twitter has allowed the Defense Department to use the social media platform to carry out a covert online propaganda and influence campaign for at least the last five years, the latest “Twitter Files” revealed Tuesday.
At the behest of the US military, the social media giant verified and “whitelisted” accounts affiliated with US Central Command in a bid to shape public opinion in countries including Yemen, Syria, Iraq, and Kuwait.
The campaign was uncovered by reporter Lee Fang of The Intercept, who said that he was allowed to make unrestricted requests for internal Twitter documents over three days last week.
Twitter executives have long claimed that the company targets and deletes secretly operated government accounts that try to spread favorable geopolitical narratives.
The US-backed accounts, which shared government-generated news items and memes, initially were openly affiliated with Washington and tweeted in Arabic. At some point, however, the Pentagon concealed the accounts’ true affiliations.
However, Twitter went against stated company policy and allowed the accounts to post messages rather than zapping them.
The first known reference to the campaign appears in a July 2017 email from a CENTCOM official to a Twitter representative requesting approval for the verification of one account and the whitelisting of 52 Arab-language accounts that the official said were used to “amplify certain messages.”
“We’ve got some accounts that are not indexing on hashtags — perhaps they were flagged as bots,” wrote Nathaniel Kahler, the CENTCOM official. “A few of these had built a real following and we hope to salvage.”
A former Twitter employee explained to the Intercept that military-run accounts were frequently marked as spam because they engaged with extremist groups, such as the Islamic State and other terror organizations.
One of the accounts flagged by Kahler was used to broadcast announcements about US drone strikes in Yemen, and another promoted US-backed militias in Syria’s long-running civil war. Still, others spread anti-Iran messages in Iraq and discussed legal issues in Kuwait.
The accounts flagged by the CENTCOM official were given a special exemption tag which functioned as a verification mechanism without a visible “blue check,” according to The Intercept.
A Twitter engineer interviewed by The Intercept said that he had never seen this type of “whitelist” tag before, but that it appeared to give accounts the benefits of Twitter verification without a visible blue check. Benefits include invulnerability to bots that flag accounts for spam or abuse, which decreases the visibility of tweets and may lead to the suspension of the account.
One of the accounts Kahler flagged to Twitter for whitelisting, @mktashif, was identified earlier this year by researchers affiliated with the Stanford Internet Observatory as one of thousands of accounts suspected of being part of a state-run influence operation that used human faces generated by artificial intelligence.
The @mktashif account initially disclosed that it was affiliated with CENTCOM but the disclosure was deleted at some point and the profile photo changed to an artificially generated face. The bio for the account translated from Arabic read, “dedicated to serving Iraqis and Arabs,” according to the Intercept, and it frequently posted tweets denouncing Iran, Houthi rebels in Yemen, and other U.S. adversaries before it was suspended earlier this year.
“It sounds like DOD was doing something shady and definitely not in line with what they had presented to us at the time,” a former Twitter employee told the Intercept.
Emails from 2020 show that Twitter and Facebook execs were invited by Pentagon officials to attend classified briefings in a sensitive compartmented information facility, or a SCIF, used for only the most highly sensitive meetings. It appears the meetings may have involved discussions on ways for the government to hide its “inauthentic activity” on social media.
“Facebook have had a series of 1:1 conversations between their senior legal leadership and DOD’s [general counsel] re: inauthentic activity,” wrote former Twitter head of trust and safety Yoel Roth, in an email obtained by the Intercept. “Per FB,” continued Roth, “DOD have indicated a strong desire to work with us to remove the activity — but are now refusing to discuss additional details or steps outside of a classified conversation.”
Stacia Cardille, a former attorney with Twitter, noted in another email that the Pentagon may want to label its clandestine social media activities as classified “to obfuscate their activity in this space, and that this may represent an overclassification to avoid embarrassment.”
Former Twitter deputy general counsel Jim Baker blasted the government in an email, writing that the Pentagon appeared to have used “poor tradecraft” in setting up the web of accounts. Baker guessed that the secret meetings were being organized because “DoD might want to give us a timetable for shutting them down in a more prolonged way that will not compromise any ongoing operations or reveal their connections to DoD.”
Lisa Roman, a top Twitter public policy official, also noted in a May 2020 email to a Pentagon lawyer that accounts that had not been explicitly provided to Twitter by the Pentagon – “may violate our Rules.” An attached list of accounts tweeted in both Russian and Arabic about human rights violations committed by ISIS, while not disclosing their US-government affiliation, according to the Intercept.
Clandestine psychological operations, such as the practice of running fake social media accounts pumping out propaganda, were legally affirmed in Section 1631 of the 2019 National Defense Authorization Act in an effort to counter disinformation campaigns by Russia, China, and other foreign threats.