OpenAI Takes Down ChatGPT Accounts Linked to Iranian Threat Actors

• OpenAI recently took down the accounts of an Iranian group that was using ChatGPT to generate content to influence global issues and the upcoming US election.
• The group is named Storm-2035 and has been active since 2020.
• They were generating both short and long-format content and posting them on websites and social media platforms.

On Friday, OpenAI said it has taken down a few accounts linked to an Iranian group that was using ChatGPT to create content to meddle with the upcoming US elections, the conflict between Gaza and Israel, and a few other issues.

The group was identified as Storm-2035 and is said to have been operational since 2020.

An investigation by Microsoft revealed that it was creating both long-format articles and short comments for social media in English and Spanish.

Common topics included the following:

• Commentary on the candidates on both sides in the US elections
• The presence of Israel in the Olympics amidst the war
• Political conditions in Venezuela
• Scottish independence

The group also had 5 websites of its own namely EvenPolitics, Nio Thinker, Savannah Time, Teorator, and Westland Sun. This is where they post the content generated from ChatGPT.

Along with ChatGPT, these websites also plagiarize content from other valid US publications. Some of these websites are progressive while some represent conservative media outlets. This way, they could target people from both ends of the political spectrum.

As for social media, they ran several accounts on X and Instagram. In order to appear more authentic, the group interspersed their political content with comments about fashion and beauty" as per OpenAI officials.

The only silver lining here is that the content they posted did not get much engagement, hardly any likes and comments.

The Danger Is Far from Over

Although OpenAI banning these accounts is a positive step, the danger of the upcoming US elections is far from over.

Microsoft had previously reported about Iran's influence on the US elections. It also said that there's been a sharp increase in foreign influence activity targeting the U.S. election over the past six months.

Along with Iranian groups, many Russian groups such as Ruza Flood (aka Doppelganger), Storm-1516, and Storm-1841 (aka Rybar) have been found spreading misinformation.

And Microsoft isn't the only one to say so.

• Meta has already disrupted 6 new operations from Russia, one from Vietnam, and one from within the US in the second quarter of 2024 alone.
• OpenAI also disrupted 5 other covert influence operations in May 2024.

Similarly, Google's Threat Analysis Group (TAG) also said that it detected and disrupted Iranian-backed spear-phishing attacks that were trying to compromise the personal accounts of high-profile users in Israel and the U.S., including those who are associated with the upcoming US elections.

Despite multiple attempts from the authorities and companies to curb such activities, these notorious groups are finding new ways to evade detection which includes spoofing well-known publications such as The New Yorker, Cosmopolitan, or Entertainment Weekly.We can't see an end to this, at least not yet.

The post OpenAI Takes Down ChatGPT Accounts Linked to Iranian Threat Actors appeared first on The Tech Report.