OpenAI reported banning two AI-enabled false-front operations, one from Russia and one from Iran, that used AI to support sophisticated covert influence campaigns. The operations used AI to generate long-form articles and social media content, aiming to spread geopolitical messaging. These operations were among the most complex and impactful the company has disrupted.
The Iranian operation used seven journalist personas to pitch long-form articles to global media outlets, while the Russian operation co-opted individuals in Latin America to run a fake think tank. Both operations used AI to draft internal reports, with the Russian operation relying heavily on AI to generate fake documents and audio scripts.
OpenAI assessed the Russian operation as a Category 5 on the IO Breakout Scale, the highest level of impact, and the Iranian operation as Category 4. Both operations managed to land their content in mainstream media, rather than just social media, indicating a significant reach and potential influence.
"Both operations closely resembled complex influence operations of the pre-AI age, but used AI to make some of the workflows easier," said OpenAI. This highlights how AI can be used to enhance the scale and sophistication of false-front operations.
The covert nature of these operations makes them vulnerable to responsible disclosure. Both "Alice Donovan" and "PeaceData" ceased their activity after being exposed, which underscores the importance of transparency in countering such threats.
OpenAI did not specify the exact number of fake stories spread by the Russian operation, and noted the challenge of tracking the full extent of these operations' impact. The company plans to continue exposing and disrupting such activities to strengthen defenses against AI-enabled false-front operations.
Source: openai
