OpenAI Blocked Two Covert Influence Operations
The firm disrupted AI-assisted campaigns, including one using a fake think tank to target users in Latin America.
Updated on Oct. 9, 2026 in Artificial Intelligence
Live Poll
Do you trust AI companies to effectively prevent the use of their tools for spreading misinformation?
OpenAI terminated access for two groups conducting covert influence operations that used AI models to create fake identities and generate content. One of the campaigns, known as Dark Clark, used a fabricated think tank called the Social Research Center to spread misinformation across Latin America.
Why it matters
The operations exploited AI to generate content and circulate it through real media channels, representing a significant test of platform safety. OpenAI classified the Dark Clark campaign as Level 5 on its IO Breakout Scale, marking the first time any operation has reached that severity.
OpenAI blocked two influence operations that leveraged generative models to create fake personas and content. The Dark Clark campaign is the first to reach Level 5 on the company's proprietary IO Breakout Scale.
The players
OpenAI
This artificial intelligence research organization develops generative models and monitors their usage for potential security threats.
The details
The groups utilized AI tools to generate sophisticated deceptive content, which was then pushed through legitimate media outlets to gain credibility. By masquerading as the Social Research Center, the Dark Clark operation sought to influence public opinion throughout Latin America.
Timeline
October 8, 2026: OpenAI disclosed the termination of the two influence operations.
The Tech Race
This detection event tracks the evolution of AI-assisted deceptive campaigns as categorized by OpenAI's IO Breakout Scale. It marks a significant shift in the arms race between safety researchers and malicious actors using generative tools to mimic authentic communication.
Users of digital news media should remain vigilant regarding the authenticity of think tanks and news sources cited online. The rise of AI-generated misinformation makes it increasingly important for readers to verify the origins of content targeting regional audiences.
The takeaway
The successful identification of the Dark Clark campaign demonstrates that AI developers are increasing their monitoring of sophisticated influence operations. Vigilance from both platform providers and the public remains essential to counter the use of synthetic media in geopolitical campaigns.
Further reading
For more context on current research and safety disclosures, visit the Artificial Intelligence section.
Live Poll
Do you trust AI companies to effectively prevent the use of their tools for spreading misinformation?
