How the UAE and MEK weaponize AI to spread propaganda
Somewhere in Iran, a group of activists used the messaging app Telegram to discuss politics and news with someone they knew as “Parsa.” They had no idea that they were actually talking to an AI agent. The system had learned to copy the real activist’s writing style by studying 8,400 of their actual Telegram posts.
Instead of talking to “Parsa,” the activists were communicating with the People’s Mojahedin Organization of Iran (MEK), an Iranian opposition group that combines Islamist and Marxist ideas and is widely described as a cult.
According to a report released by Anthropic last month, four people working for the MEK carried out the operation. They built an AI agent called “Viktor,” which also watched people inside Iran and filled social media platforms with pro-MEK messages. The accounts made their content look as if it came from independent sources.
The Anthropic report shows how foreign groups are increasingly using artificial intelligence tools such as Claude and ChatGPT to influence what people think. AI agents like Viktor are computer systems powered by large language models that can carry out tasks on their own. Many of the activities described in the report, such as writing reports, creating content, and collecting information, look much like the everyday tasks that millions of Americans use Claude and ChatGPT to do.
One of the report’s biggest surprises is how widely groups and partners aligned with the United States have also begun using AI to influence public opinion.
The MEK, for example, has important supporters in the U.S. Congress. Last year, 230 members of the House of Representatives co-sponsored a resolution calling for U.S. support for the MEK’s plan for Iran. Former National Security Adviser John Bolton, former Vice President Mike Pence, and former New York City Mayor Rudy Giuliani are among the well-known Republican politicians who have spoken at MEK rallies, often for a large speaking fee.
One Instagram account linked to the MEK’s Viktor system, @Tehranchekhabar19, has more than 170,000 followers. It presents itself as an independent news account but publishes material that supports the MEK’s views. This includes videos that support the U.S.-Israeli war against Iran and criticize Reza Pahlavi, Iran’s former crown prince, who leads a rival opposition group. Some of the account’s videos attract hundreds of thousands of views. Viktor also created other Instagram accounts with hundreds of thousands of followers. Some appear to have been shut down since then.
The United States listed the MEK as a Foreign Terrorist Organization until 2012. The group was even more active on Telegram, where it tried to secretly recruit people inside Iran to support its goals. According to Anthropic, Viktor entered more than 500 Telegram groups. It then “grouped these targets by city, age, occupation, political alignment, and arrest history likely to help them tailor their messages to the specific audiences.”
Craig Holman, a government affairs lobbyist for Public Citizen, told Responsible Statecraft by email that this was a new kind of influence tactic. “These accounts of foreign governments using AI to manipulate the opinions of legislative bodies and the public are based on the root source of AI deception: falsely depicting a real person or entity doing or saying something that they never did.”
This case and others described in similar reports show that foreign groups are taking practical steps to use AI as a weapon in information wars. For example, a report published by OpenAI last week described how Iranian operators used ChatGPT to place almost 100 articles in publications covering international affairs. The articles appeared under fake author names. They generally argued against war with Iran, with headlines such as “Is America’s Iran Policy Being Privatized? The Jared Kushner Question” and “Trump’s Iran Gamble Hardened the Regime It Meant to Break.”
The recent OpenAI report also found that Russia had created and run a fake think tank called the “Social Research Center.” It employed real staff in Latin America who did not know they were working for Russia. Russian operators used ChatGPT to create content and write internal reports. Meanwhile, the staff, unaware of who was behind the organization, conducted real interviews with experts and published research papers, mostly about relations between India and Latin America.
The Social Research Center’s account on X has been suspended, but its website was still active when the report was published. The website carried a statement rejecting OpenAI’s claims as “factually incorrect and as constituting a deliberate attempt to discredit an independent research project.”
Another case described how one of Washington’s closest partners in the Middle East is using AI to try to hide its role in a bloody civil war.
The UAE’s “Deadshot”
An example of posts published by accounts supporting the UAE about Sudan’s civil war. (Screenshot via Anthropic.com)
Officials from the United Arab Emirates (UAE) created their own AI agent using Claude. They called it “Deadshot” and gave it a major task: to make the UAE’s role in Sudan’s civil war look better. A Sudanese armed group called the Rapid Support Forces (RSF), which is backed by the UAE, has been credibly accused of ethnic cleansing and sexual violence. A United Nations fact-finding mission had said that the RSF’s actions against non-Arab communities in Sudan’s Darfur region showed signs of genocide.
The AI agent quickly began working to protect the UAE’s reputation, especially at the United Nations. According to Anthropic’s report, Deadshot took on the identity of a real Sudanese human rights organization and secretly wrote testimony for the U.N. The aim was to influence the organization by making the testimony appear to come from “independent local witnesses, rather than state messaging.”
Anthropic’s report gives conflicting information about whether this effort succeeded. In one section, it says the company could not confirm whether the testimony reached the people it was meant for. Elsewhere, a summary says the testimony was “delivered in a live UN Human Rights Council session.” Anthropic did not respond to a request for comment.
Deadshot also collected information about people who criticized the UAE. It prepared what the report called “counter-accountability dossiers” on U.N. Special Rapporteurs who had spoken out against the UAE’s role in Sudan. These officials investigate and report on human rights issues.
The UAE also ran a wider campaign to influence public opinion, mainly on X. Deadshot created and managed about 300 fake or misleading social media accounts. They posted messages blaming the Sudanese civil war on the Muslim Brotherhood, which supports the Sudanese Armed Forces in their fight against the RSF.
Many of the posts made appeals for people affected by the humanitarian crisis. But they shifted blame away from the RSF and toward the Sudanese Muslim Brotherhood (SMB), which the United States designated as a Foreign Terrorist Organization earlier this year. One post on X, published by an account with 27,000 followers that Anthropic linked to the campaign, said: “The longer Sudan’s war drags on, the deeper the famine crisis becomes. Analysts point to hardline Islamist elements supporting continued military confrontation over negotiated settlements.” Another post from an account with 119,000 followers said the SMB’s role in the conflict “makes ending the war far harder.”
According to the report, the operator told Claude to carry out a “coordinated transatlantic and regional operation to dismantle the Muslim Brotherhood globally” hundreds of times.
Kholood Khair, founding director of Confluence Advisory, a think tank based in Khartoum, told Responsible Statecraft that Sudan’s conflict is being fought on three fronts: the battlefield, negotiations to end the war, and the fight to shape public opinion. She said the information campaign is becoming more important. “These posts are carefully deployed to attempt to co-opt the pro-civilian rule, pro-democracy movement and camouflage what the UAE is actually trying to achieve in Sudan,” Khair said. She explained that many posts do not contain facts that are clearly false, but they leave out information and tell only part of the story.
“AI is the next battleground in many of these wars, and Sudan’s war has clearly become a testing ground, with the UAE getting a head start,” Khair added.