OpenAI has disrupted several covert influence operations using its AI technology to manipulate public opinion globally. These operations targeted political outcomes and spread misinformation through social media and other online platforms.

OpenAI, the artificial intelligence firm founded by Sam Altman, has recently identified and terminated accounts linked to several covert influence operations that exploited its technology to manipulate public opinion worldwide. The firm revealed these actions on May 30, stating that it had dismantled five such operations in the past three months.


Disruption of Covert Influence Operations

OpenAI's models were misused by bad actors to generate comments for articles, create social media personas, and handle translations and proofreading of texts. The firm highlighted that these influence operations aimed to deceive and manipulate audiences across various platforms.


Operation "Spamouflage"

One significant operation, named "Spamouflage," utilized OpenAI's technology to conduct social media research and create multilingual content on platforms like X, Medium, and Blogspot. The content was designed to influence public opinion and political outcomes. Additionally, Spamouflage used AI to debug code and manage databases and websites.


Operation "Bad Grammar"

Another operation, called "Bad Grammar," targeted regions including Ukraine, Moldova, the Baltic States, and the United States. This operation used OpenAI models to run Telegram bots and generate political comments, furthering its deceptive activities.


Operation "Doppelganger"

The "Doppelganger" group used AI models to produce comments in multiple languages, including English, French, German, Italian, and Polish. These comments were posted on platforms like X and 9GAG to manipulate public opinion.


International Union of Virtual Media

The "International Union of Virtual Media" leveraged AI to generate long-form articles, headlines, and website content, which were then published on their affiliated websites. This operation aimed to disseminate misleading information through credible-looking content.


Commercial Operation STOIC

OpenAI also disrupted a commercial entity known as STOIC. This company used AI to create articles and comments for social media platforms such as Instagram, Facebook, X, and other websites associated with its operations.


Scope of Misinformation

The various operations focused on a wide array of issues, including:

  • Russia’s invasion of Ukraine

  • The conflict in Gaza

  • Indian elections

  • Politics in Europe and the United States

  • Criticisms of the Chinese government by dissidents and foreign governments


Ben Nimmo, a principal investigator for OpenAI, elaborated on these findings in a report shared with The New York Times. He highlighted that the case studies exemplify some of the most prominent and persistent influence campaigns currently active.


Impact and Implications

The New York Times noted that this is the first instance where a major AI firm has disclosed how its specific tools were used for online deception. OpenAI concluded that, to date, these operations have not achieved significantly increased audience engagement or reach through the use of its services.


OpenAI's actions underscore the ongoing challenges in combating misinformation and manipulation in the digital age. By identifying and terminating these covert influence operations, OpenAI aims to protect the integrity of online discourse and ensure that its technology is used ethically and responsibly. The firm's proactive measures highlight the importance of vigilance and transparency in the AI industry to safeguard public trust and democratic processes.


(MARTIN YOUNG, COINTELEGRAPH, 2024)