According to a recent Microsoft report, Iran is increasingly relying on cyber-based influence operations and generative AI to interfere with the 2024 US presidential elections, while Russia and China continue to pose active threats.
Since June 2024, Iran has been preparing for influence operations targeting US entities, including initial cyber reconnaissance and the infiltration of online personas and websites.
Microsoft has identified several Iranian groups involved in these activities. The Sefid Flood group has been preparing for influence operations since the Iranian New Year’s festival in late March.
They specialize in impersonating social and political activists to sow chaos and undermine trust in authorities.
Another group called Mint Sandstorm, linked to Iran’s Revolutionary Guards, attempted to phish a senior staffer of a presidential campaign in June. The group had carried out similar attacks before the 2020 election.
Microsoft reports on an Iranian network called Storm-2035 that operates four websites posing as news portals. These target US voter groups across the political spectrum, disseminating polarizing messages on topics like presidential candidates, LGBTQ rights, or the Ukraine war.
The report states that Iranian actors are using AI-based services to plagiarize and rephrase content from US publications. They employ SEO plugins and other AI-based tools to generate article headlines and keywords.
In addition, Russian and Chinese actors remain active. Microsoft is observing three Russian influencers running campaigns related to the US elections. The Storm-1516 group has been focusing on producing fake videos spreading scandalous claims since April.
Chinese actors such as Taizi Flood and Storm-1852 have broadened their presence to different platforms and adapted their strategies.
They utilize numerous accounts to provoke controversy during pro-Palestinian protests at American universities. There’s a growing shift towards creating short videos discussing political issues.
Microsoft emphasizes that the actors’ use of generative AI has had minimal impact so far, and many are reverting to well-established methods: “Nearly all actors have attempted to incorporate AI content in their operations, but more recently many have shifted back to techniques that have been effective before—simple digital alterations, misrepresentation of content, and use of trusted labels or logos on false information.”
Stories You May Like