OpenAI Disrupts Covert Influence Operations Originating from Russia and Iran
Two separate covert operations that misused artificial intelligence models to spread geopolitical messages to target audiences have been uncovered.
OpenAI has announced that two separate covert influence operations based in Russia and Iran used artificial intelligence models to create fake front organizations and spread geopolitical messages to target audiences.
Disruption of Covert Operations
OpenAI announced that it has banned two covert influence operations that used artificial intelligence models for malicious purposes to support sophisticated fake front formations and launder geopolitical content.
Iran-Origin Fake Journalist Activities
In the Iran-origin operation, it was discovered that seven different fake journalist identities were used to propose long articles to global media organizations and generate social media comments.
Russia-Linked Dark Clark Operation
The Russia-origin operation, known by the alias Dark Clark, operated a think tank called the Center for Social Research using unsuspecting individuals in Latin America and produced fake leaked documents.
Impact Scale Levels of the Operations
According to OpenAI's Impact Scale system, the Russia-origin activity reached category 5, becoming the first such operation blocked by the company, while the Iran-origin activity reached category 4.