OpenAI has suspended ChatGPT accounts linked to an Iranian crew that may have spread misleading material about the upcoming US presidential election campaign on social media.
ChatGPT accounts linked to an Iranian crew that may have been spreading false information about the upcoming US presidential election campaign on social media have been suspended by OpenAI. As we were told, the accounts used ChatGPT to compose short comments and longer posts to post on social media about Democratic presidential candidate Kamala Harris and Republican nominee Donald Trump for the US presidency. Although the influence operation was posted on many websites, it “does not appear to achieve significant audience engagement,” according to an alert issued Friday.
Earlier this week, Google’s threat hunters released information on Iranian cyber influence activities in response to a spike in attacks that resulted in data leaks from the Trump re-election campaign. In total, the AI organization said it identified 12 accounts on X and one on Instagram involved in this covert influence operation. In addition to the US presidential election, the fake news and commentary covered the conflict in Gaza and Israel’s participation in the Olympics, among other topics. OpenAI attributed the fake posts to Storm-2035, a Tehran-backed group that Microsoft also raised concerns about last week.
Using five websites that mimicked both conservative and progressive news sources, the first step of the work produced articles about U.S. politics and world events, OpenAI said in a statement released Friday. “The second step of the work produced short commentaries in English and Spanish that were posted on social media.”
These five areas are:
niothinker(.)com
SavannahTime (.)evenpolitics.com(.)
Westlandsun(.)com com theorator(.)com
In the run-up to the 2022 U.S. midterm elections, Storm-2035 and at least one of its fake news sites, EvenPolitics, have been active online, publishing about ten “articles” per week. While these are written in English, the group also maintains additional influence operations websites in Arabic, English, French, and Spanish.
Storm-2035 received only a 2 on the Brookings Breakout Scale, which rates the effectiveness of such covert operations on a scale of 1 (lowest) to 6 (highest). This indicates that while the malware was active on multiple platforms, it showed no real signs of being discovered by real people or widely distributed.
“The majority of the social media posts we identified received few or no likes, shares or comments,” OpenAI noted. “We also found no evidence that the web articles were shared on social media.” Five accounts that OpenAI claimed were used by state-backed organizations in China, Iran, Russia and North Korea to create dangerous software scripts and phishing emails were shut down earlier this year.
Read also: At Jar, we have leveraged cutting-edge technology to improve the efficiency and usability of our platform, says Nishchay Ag, co-founder and CEO of Jar
Follow: CIO News LinkedIn Account | CIO News Facebook | CIO News Youtube | CIO News on Twitter
About Us:
CIO News is the leading platform for delivering the latest news, updates and insights from the CIO industry. As a trusted source in the technology and IT sector, we provide a comprehensive resource for executives and professionals who want to stay informed and up to date. With a focus on cutting-edge developments and trends, CIO News is your go-to source for staying up to date with the rapidly evolving technology and IT landscape. CIO News was founded in June 2020 and has grown rapidly, with ambitious growth plans for global expansion with a focus on markets in the Middle East and Africa, ASEAN, the US and the UK.
CIO News is owned by Mercadeo Multiventures Pvt Ltd.
