- OpenAI intensifies its crackdown on Iran-linked accounts spreading US election misinformation using ChatGPT.
- OpenAI’s investigation reveals Iran’s AI-driven influence operation, targeting both progressive and conservative audiences.
- The crackdown highlights OpenAI’s commitment to combating foreign interference in US elections via AI technology.
OpenAI has intensified its crackdown on Iran-linked accounts exploiting its platforms to spread misinformation about the US Presidential election. The recent surge in these activities has garnered significant attention, especially after Donald Trump’s election campaign reported a security breach linked to Iranian hackers. The AI firm has identified and banned numerous accounts associated with Iran’s influence operation, dubbed “Storm-2035.”
Targeted Misinformation Across Various Topics
The accounts used OpenAI’s generative AI technologies, particularly ChatGPT, to create and disseminate false information on various topics, with a strong focus on the US election. These operations involved producing both long-form articles and brief social media comments, shared across multiple platforms. The topics ranged from the US election to the Gaza conflict, Israel’s participation in the Olympic Games, and others.
Moreover, OpenAI classified the operation as a low-level threat based on Brookings’ Breakout Scale, noting that the posts received little engagement. Despite the limited impact, the firm recognized the importance of monitoring and preventing such foreign influence operations, emphasizing the need for heightened vigilance.
OpenAI’s Broader Commitment to AI Ethics
Significantly, the AI firm’s investigation revealed that the operation targeted both progressive and conservative audiences. The accounts posed as legitimate news outlets and social media users, sometimes mimicking real users’ comments to appear authentic. This tactic underscored the sophisticated nature of the operation, despite its limited success.
Besides, the timing of OpenAI’s announcement is notable, coming shortly after Trump’s campaign reported a security breach attributed to Iranian hackers. The company’s actions highlight its broader commitment to transparency and the ethical use of AI, particularly in safeguarding political processes from foreign interference.
Ben Nimmo, OpenAI’s principal investigator for intelligence and investigations, stated that this was the first instance where the company detected an operation specifically targeting the US election.
He stressed the importance of staying alert but maintaining calm, as the operation had minimal reach. Additionally, the report contributes to the growing evidence of Iran’s tech-centric attempts to influence the US election.
