OpenAI has uncovered an Iranian group using ChatGPT to create fake news articles to try and influence US voters during the presidential election. “This week we identified and took down a cluster of ChatGPT accounts that were generating content for a covert Iranian influence operation,” the company said in a blog post on Friday. OpenAI shut down the accounts after company partner Microsoft flagged a group of news sites that it suspects have been the work of an Iranian propaganda outfit, dubbed “Storm-2035.”The sites, which include NioThinker and Savannah Time, claim to cater to left-wing and conservative readers, respectively. However, Microsoft saw signs that both media outlets are using generative AI to pump out articles in an effort to sway American voters. OpenAI has now confirmed that the Iranian group was using ChatGPT “to generate content focused on a number of topics—including commentary on candidates on both sides in the US presidential election—which it then shared via social media accounts and websites.” Other sites allegedly linked to the Iranian group include Even Politics, Teorator, and Westland Sun.
(Credit: OpenAI)
Although the ChatGPT-created articles were successfully posted on the Iranian group’s websites and social media accounts, OpenAI said: “We have seen no indication that this content reached a meaningful audience.”“The majority of social media posts that we identified received few or no likes, shares, or comments. We similarly did not find indications of the web articles being shared across social media,” the company added.
Recommended by Our Editors
(Credit: OpenAI)
The incident underscores how state-sponsored groups can abuse today’s chatbots to create authentic-sounding propaganda for American audiences. In a related case, back in February, OpenAI uncovered at least five state-sponsored outfits from Iran, China, Russia and North Korea all using generative AI to refine their hacking techniques. As for Storm-2035, the Iranian group was tapped ChatGPT both to create long-form articles and posts for social media. The exact aim behind the influence operation isn’t entirely clear since it appears the AI-generated posts have been critical of both Trump and Democrats. But in addition to the US election, other topics included the conflict in Gaza, Israel’s presence at the Olympic Games, and politics in Venezuela, among others. OpenAI has since banned the Iranian-linked accounts, although there’s always a chance the Iranian group will find new ways to abuse ChatGPT. OpenAI says it remains on the watch “for any further attempts to violate our policies.” The company has also shared more details about the threat to the US government, the campaign races, and “industry stakeholders,” it said.
Get Our Best Stories!
Sign up for What’s New Now to get our top stories delivered to your inbox every morning.
This newsletter may contain advertising, deals, or affiliate links. Subscribing to a newsletter indicates your consent to our Terms of Use and Privacy Policy. You may unsubscribe from the newsletters at any time.
About Michael Kan
Senior Reporter
I’ve been with PCMag since October 2017, covering a wide range of topics, including consumer electronics, cybersecurity, social media, networking, and gaming. Prior to working at PCMag, I was a foreign correspondent in Beijing for over five years, covering the tech scene in Asia.
Read Michael’s full bio
Read the latest from Michael Kan
